Skip to content

I
imagej-elphel
Commit b6994b6e authored by Andrey Filippov's avatar Andrey Filippov
Browse files
Options

Fixing field correction

parent 3868b0a7
Hide whitespace changes
Inline Side-by-side
Showing with 576 additions and 130 deletions
src/main/java/AlignmentCorrection.java
View file @ b6994b6e
......@@ -32,6 +32,14 @@ public class AlignmentCorrection {
static final int NUM_SLICES = 10; // disp, strength, dx0, dy0, dx1, dy1, dx2, dy2, dx3, dy3)
static final int NUM_ALL_SLICES = 14; // disp, strength, dx0, dy0, str0, dx1, dy1, str1, dx2, dy2, str2, dx3, dy3 str3,)
static final int NUM_SENSORS = 4;
static final int [] INDICES_14_10 = {0,1,2,3,5,6,8,9,11,12};
static final int [] INDICES_10_DISP = {2,4,7,9}; // which indices need to add disparity to restore full offsets
static final int [] INDICES_14_WEIGHTS = {4,7,10,13}; // now pair weights can be zeros where common weight is not (and respective values are NaNs)
static final int INDEX_14_WEIGHT = 1;
static final int INDEX_10_WEIGHT = 1;
static final int INDEX_10_DISPARITY = 0;
static final int INDEX_14_DISPARITY = 0;
QuadCLT qc;
public class Sample{
......@@ -72,10 +80,25 @@ public class AlignmentCorrection {
{
this.use_disparity = use_disparity;
this.pXY = pXY;
//FIXME: 2 next fields are not used
this.disparity_task = disparity_task; // currently wrong, series is 0/1, not measurement number
this.disparity_meas = disparity_meas;
this.strength = strength;
this.strength = (Double.isNaN(strength) || (offsets==null) || Double.isNaN(pXY[0]) || Double.isNaN(pXY[1]))?0.0:strength;
this.offsets = offsets;
if (strength != 0.0) {
boolean bug = (offsets == null);
if (!bug) {
for (int i = 0; i < offsets.length; i++) {
bug |= (offsets[i] == null) || Double.isNaN(offsets[i][0]) || Double.isNaN(offsets[i][1]);
if (bug) break;
}
}
if (bug) {
System.out.println("***** BUG: NaN offsets, but strength != 0: x= "+pXY[0]+", y = "+pXY[1]+" ******");
System.out.println("***** BUG: NaN offsets, but strength != 0: x= "+pXY[0]+", y = "+pXY[1]+" ******");
this.strength = 0;
}
}
}
public boolean usesDisparity()
......@@ -154,15 +177,20 @@ public class AlignmentCorrection {
double [] w,
int n_sample)
{
// for (int i = n_sample * (2 * NUM_SENSORS); i < (n_sample + 1) * (2 * NUM_SENSORS); i++){
for (int i = 0; i < 2 * NUM_SENSORS; i++){
w[n_sample * (2 * NUM_SENSORS) + i] = (use_disparity || (i < 7)) ? strength : 0.0;
// w[n_sample * (2 * NUM_SENSORS) + i] = (!Double.isNaN(strength) && (use_disparity || (i < 7))) ? strength : 0.0;
// TODO: find out about (i < 7) - it zeroes out y[3]?
w[n_sample * (2 * NUM_SENSORS) + i] =(use_disparity || (i < 7)) ? strength : 0.0; // Last element corresponds to disparity,
// set weight = 0 for samples that do not have it.
}
}
// measurement vectors and correction vectors - seem to mismatch for disparity (sym0)
/*
public double [][] get_dMismatch_dXY()
{
double [][] dMismatch_dXY = { // extra 0.5 is because differences dxi, dyi are already *= 0.5/magic
// FIXME: 0.5/magic is removed, but what is measured is (x1-x0)/2m ...
//x0 y0 x1 y1 x2 y2 x3 y3
{ 0.0 , -0.5, 0.0 , 0.5 , 0.0 , 0.0 , 0.0 , 0.0 }, // mv0 = dy0 = y1 - y0
{ 0.0 , 0.0 , 0.0 , 0.0 , 0.0 , -0.5 , 0.0 , 0.5 }, // mv1 = dy1 = y3 - y2
......@@ -174,6 +202,24 @@ public class AlignmentCorrection {
{-0.0625, -0.0625, 0.0625,-0.0625,-0.0625, 0.0625, 0.0625, 0.0625}};// mv7 = (dx0 + dx1 +dy2 + dy3)/8= (x1 - x0 + x3 - x2 + y2 - y0 + y3 - y1)/8
return dMismatch_dXY;
}
*/
public double [][] get_dMismatch_dXY()
{
double [][] dMismatch_dXY = { // extra 0.5 is because differences dxi, dyi are already *= 0.5/magic
// FIXME: 0.5/magic is removed, but what is measured is (x1-x0)/2m ...
//x0 y0 x1 y1 x2 y2 x3 y3
{ 0.0 , -0.5, 0.0 , 0.5 , 0.0 , 0.0 , 0.0 , 0.0 }, // mv0 = dy0 = y1 - y0
{ 0.0 , 0.0 , 0.0 , 0.0 , 0.0 , -0.5 , 0.0 , 0.5 }, // mv1 = dy1 = y3 - y2
{-0.5 , 0.0 , 0.0 , 0.0 , 0.5 , 0.0 , 0.0 , 0.0 }, // mv2 = dx2 = x2 - x0
{ 0.0 , 0.0 , -0.5 , 0.0 , 0.0 , 0.0 , 0.5 , 0.0 }, // mv3 = dx3 = x3 - x1
{ 0.25, 0.0 , -0.25, 0.0 , -0.25, 0.0 , 0.25, 0.0 }, // mv4 = (dx1 - dx0)/2 = (x3 - x2 + x0 - x1) / 2
{ 0.0 , 0.25, 0.0 , -0.25, 0.0 , -0.25, 0.0 , 0.25 }, // mv5 = (dy3 - dy2)/2 = (y3 - y1 + y0 - y2) / 2
{-0.125, 0.125, 0.125, 0.125, -0.125, -0.125, 0.125, -0.125 }, // mv6 = (dx0 + dx1 -dy2 - dy3)/4 = (x1 - x0 + x3 - x2 - y2 + y0 - y3 + y1)/4
{-0.0625, -0.0625, 0.0625,-0.0625,-0.0625, 0.0625, 0.0625, 0.0625}};// mv7 = (dx0 + dx1 +dy2 + dy3)/8= (x1 - x0 + x3 - x2 + y2 - y0 + y3 - y1)/8
return dMismatch_dXY;
}
/**
......@@ -181,7 +227,7 @@ public class AlignmentCorrection {
* where sum of measurement vectors squared is minimized. Same matrix multiplications
* is applied to each group of 8 columns. last column in each group is only non-zero if
* disparity is known to be 0;
* @param jt transposed Jacobian of 10/9 rows and 8*n columns
* @param jt transposed Jacobian of 13/10/9 rows and 8*n columns
* @return converted transposed Jacobian of the same dimensions
*/
double [][] convertJt_mv(
......@@ -252,25 +298,43 @@ public class AlignmentCorrection {
tilesX);
min_strength = 0; // all > 0
if (debugLevel > 0){
String [] titles= {"disp","strength","dx0","dy0","dx1","dy1","dx2","dy2","dx3","dy3"};
double [][] dbg_img = disp_strength.clone();
for (int n = 0; n < disp_strength.length; n++){
dbg_img[n] = disp_strength[n].clone();
}
for (int n = 0; n < dbg_img.length; n+=2){
for (int n = 0; n < dbg_img.length; n++) if (n != 1){
for (int i = 0; i < dbg_img[n].length; i++) {
if (dbg_img[n+1][i] == 0.0){
if (dbg_img[1][i] == 0.0){
dbg_img[n][i] = Double.NaN;
}
}
}
(new showDoubleFloatArrays()).showArrays(dbg_img, tilesX, disp_strength[0].length/tilesX, true, "filtered_ds"); // , titles);
(new showDoubleFloatArrays()).showArrays(dbg_img, tilesX, disp_strength[0].length/tilesX, true, "filtered_ds", titles);
}
} else {
disp_strength = disp_strength_in;
min_strength = min_strength_in;
if (debugLevel > 0){
String [] titles= {"disp","strength","dx0","dy0","dx1","dy1","dx2","dy2","dx3","dy3"};
double [][] dbg_img = disp_strength.clone();
for (int n = 0; n < disp_strength.length; n++){
dbg_img[n] = disp_strength[n].clone();
}
for (int n = 0; n < dbg_img.length; n++) if (n != 1){
for (int i = 0; i < dbg_img[n].length; i++) {
if (dbg_img[1][i] == 0.0){
dbg_img[n][i] = Double.NaN;
}
}
}
(new showDoubleFloatArrays()).showArrays(dbg_img, tilesX, disp_strength[0].length/tilesX, true, "inp_ds" , titles);
}
}
if (hist_smpl_side > 0) { // 0 to bypass histogram filtering
disp_strength = filterHistogramFar (
......@@ -285,18 +349,27 @@ public class AlignmentCorrection {
hist_norm_center, // final boolean norm_center, // if there are more tiles that fit than minsamples, replace with
tilesX); // final int tilesX)
if (debugLevel > 0){
String [] titles= {"disp","strength","dx0","dy0","dx1","dy1","dx2","dy2","dx3","dy3"};
double [][] dbg_img = disp_strength.clone();
for (int n = 0; n < disp_strength.length; n++){
dbg_img[n] = disp_strength[n].clone();
}
for (int n = 0; n < dbg_img.length; n+=2){
// for (int n = 0; n < dbg_img.length; n+=2){
// for (int i = 0; i < dbg_img[n].length; i++) {
// if (dbg_img[n+1][i] == 0.0){
// dbg_img[n][i] = Double.NaN;
// }
// }
// }
for (int n = 0; n < dbg_img.length; n++) if (n != 1){
for (int i = 0; i < dbg_img[n].length; i++) {
if (dbg_img[n+1][i] == 0.0){
if (dbg_img[1][i] == 0.0){
dbg_img[n][i] = Double.NaN;
}
}
}
(new showDoubleFloatArrays()).showArrays(dbg_img, tilesX, disp_strength[0].length/tilesX, true, "hist_filt_ds"); // , titles);
(new showDoubleFloatArrays()).showArrays(dbg_img, tilesX, disp_strength[0].length/tilesX, true, "hist_filt_ds", titles);
}
}
......@@ -324,7 +397,9 @@ public class AlignmentCorrection {
use_poly, // final boolean use_poly,
clt_parameters.fcorr_inf_quad, // final boolean use_quadratic,
clt_parameters.fcorr_inf_vert, // final boolean use_vertical,
clt_parameters.ly_inf_en, // final boolean use_disparity, // for infinity
clt_parameters.ly_inf_en, // final boolean use_disparity, // for infinity - if true, restores differences in the direction of disparity that was subtracted during measurement)
// For ly_inf_en need to make sure that programmed disparity was 0.0, so
clt_parameters.ly_inf_disp, //final boolean allow_dispatity,
clt_parameters,
disp_strength,
samples_list,
......@@ -335,7 +410,7 @@ public class AlignmentCorrection {
if (debugLevel > -1){
System.out.println("infinityCorrection(): coefficient increments from infinityMismatchCorrection");
if (mismatch_corr_coefficients == null) { // non-null only for poly !
System.out.println("imismatch_corr_coefficients == null");
System.out.println("imismatch_corr_coefficients == null (non-null for polynomial correction only)");
return mismatch_corr_coefficients;
}
show_fine_corr(
......@@ -397,14 +472,14 @@ public class AlignmentCorrection {
final double fcorr_radius,
final boolean use_vertical,
final double min_strength,
final double max_diff,
final double max_diff, // also temporarily maximal difference from 0.0
final int max_iterations,
final double max_coeff_diff,
final double far_pull, // = 0.2; // 1; // 0.5;
EyesisCorrectionParameters.CLTParameters clt_parameters,
double [][] disp_strength,
int tilesX,
double magic_coeff, // still not understood coefficient that reduces reported disparity value. Seems to be around 8.5
double magic_coeff, // still not understood coefficient that reduces reported disparity value. Seems to be around 0.85
int debugLevel)
{
final int numTiles = disp_strength[0].length;
......@@ -416,6 +491,36 @@ public class AlignmentCorrection {
double thresholdQuad = 1.0E-30; // threshold ratio of matrix determinant to norm for quadratic approximation (det too low - fail)
double [] disp_surface = new double[numTiles];
ArrayList<Sample> samples_list = new ArrayList<Sample>();
// start with average disparity:
double sdw= 0.0, sw = 0.0;
int num_samples = 0;
for (int num_set = 0; num_set < disp_strength.length/NUM_SLICES; num_set++){
int disp_index = NUM_SLICES * num_set;
int str_index = NUM_SLICES * num_set + 1;
for (int nTile = 0; nTile < numTiles; nTile++) if (center_mask[nTile]){
if (disp_strength[str_index][nTile] > min_strength) {
//clt_parameters.fcorr_inf_diff
if (Math.abs(disp_strength[disp_index][nTile]) <= max_diff) {
double weight= disp_strength[str_index][nTile];
sdw += weight * disp_strength[disp_index][nTile];
sw += weight;
num_samples ++;
} else {
disp_strength[str_index][nTile] = 0.0;
}
}
}
}
if (sw != 0) {
sdw /= sw;
}
disparity_poly[5] = sdw;
if (debugLevel > 0) {
System.out.println("selectInfinityTiles(): Number of input samples exceeding "+min_strength+" strength is "+num_samples+", average disparity is "+sdw);
}
for (int pass = 0; pass < max_iterations; pass++){
for (int nTile = 0; nTile < numTiles; nTile++){
int tileX = nTile % tilesX;
......@@ -480,7 +585,7 @@ public class AlignmentCorrection {
}
indx ++;
}
if ((debugLevel > 2) && (pass < 20)){
if ((debugLevel > 2) && (pass < 21)){
String [] titles = {"disparity","approx","diff", "strength"};
double [][] dbg_img = new double [titles.length][numTiles];
for (int nTile = 0; nTile < numTiles; nTile++){
......@@ -586,7 +691,7 @@ public class AlignmentCorrection {
* @param debugLevel debug level
* @return per sub-camera, per direction (x,y) 6 quadratic polynomial coefficients, same format as fine_geometry_correction()
*/
public double [][][] infinityCorrection(
public double [][][] infinityCorrection_old(
// final double min_strength0,
// final double max_diff0,
// final int max_iterations0,
......@@ -763,7 +868,8 @@ public class AlignmentCorrection {
final boolean use_poly,
final boolean use_quadratic,
final boolean use_vertical,
final boolean use_disparity, // for infinity
final boolean use_disparity, // for infinity // now disabled?
final boolean allow_dispatity,
EyesisCorrectionParameters.CLTParameters clt_parameters,
double [][] disp_strength,
ArrayList<Sample> samples_list,
......@@ -772,6 +878,8 @@ public class AlignmentCorrection {
ArrayList<Mismatch> mismatch_list,
int debugLevel)
{
int dbgTileX = 100;
int dbgTileY = 100;
// Mismatch data has disparity values already subtracted, so to correct disparity at infinity, disparity values should be restored
final int num_tiles = disp_strength[0].length;
final int tilesY = num_tiles/tilesX;
......@@ -800,14 +908,19 @@ public class AlignmentCorrection {
{1.0, 2.0, 3.0}};
Matrix A = new Matrix(A_arr);
Matrix AINV = A.inverse();
double scale = 0.5/magic_coeff;
double scale = 1.0; // Why was it here? 0.5/magic_coeff;
double [][] dbg_xy = null;
if (clt_parameters.show_extrinsic && (debugLevel > -2)) { // TODO: Add clt_parameters
dbg_xy = new double [9][num_tiles];
// dbg_xy = new double [9][num_tiles];
dbg_xy = new double [10][num_tiles];
}
for (Sample s: samples_list){
int tileX = s.tile % tilesX;
int tileY = s.tile / tilesX;
if ((debugLevel > 0) && (tileX == dbgTileX) && (tileY == dbgTileY)) {
System.out.println("infinityMismatchCorrection(): tileX = "+tileX+", tileY = "+tileY);
}
double [] xy = new double[8]; // same as coefficients: x0,y0,x1,y1,x2,y2,x3,y3
// Calculate x0,x1,x2,x3 and y0,y1,y2,y3 assuming x0+x1+x2+x3 = 0,y0+y1+y2+y3 = 0 and minimizing squares of errors
// as each each 4: "dx0", "dx1", "dx2", "dx3" and "dy0", "dy1", "dy2", "dy3" are over-defined
......@@ -816,6 +929,13 @@ public class AlignmentCorrection {
// double [] dxy = new double[4];
for (int i = 0; i < 4; i++){
dxy[i][dir] = scale * disp_strength[indices_mismatch[dir][i] + (s.series * NUM_SLICES)][s.tile];
if (Double.isNaN(dxy[i][dir])) {
int ii = indices_mismatch[dir][i] + (s.series * NUM_SLICES);
System.out.println("**** BUG: infinityMismatchCorrection() tileX="+tileX+", tileY="+tileY+" dxy["+i+"]["+dir+"]= "+dxy[i][dir]+
", index="+ii+", s.tile = "+s.tile);
System.out.println("**** BUG: infinityMismatchCorrection() tileX="+tileX+", tileY="+tileY+" dxy["+i+"]["+dir+"]= "+dxy[i][dir]+
", index="+ii+", s.tile = "+s.tile);
}
}
/*
......@@ -902,7 +1022,7 @@ B = |+dy0 -dy1 -2*dy3 |
*/
double [] B_arr = {
-dxy[0][dir] -dxy[1][dir] -dxy[2][dir] -dxy[3][dir],
-dxy[0][dir] -dxy[1][dir] -dxy[2][dir] -dxy[3][dir],
dxy[0][dir] -dxy[1][dir] -2 * dxy[3][dir],
dxy[2][dir] -2 * dxy[1][dir] -dxy[3][dir]};
Matrix B = new Matrix(B_arr, 3); // 3 rows
......@@ -923,6 +1043,7 @@ B = |+dy0 -dy1 -2*dy3 |
for (int i = 0; i < xy.length; i++){
dbg_xy[i][s.tile] += xy[i] * s.weight;
}
dbg_xy[9][s.tile] += s.weight*(-xy[0]-xy[1]+xy[2]-xy[3]-xy[4]+xy[5]+xy[6]+xy[7])/8;
dbg_xy[8][s.tile] += s.weight;
}
......@@ -960,10 +1081,10 @@ B = |+dy0 -dy1 -2*dy3 |
// final int disp_scan_count,
mismatch_list.add(new Mismatch(
false, // public boolean use_disparity; // adjust dx0+dx1+dy0+dy1 == 0
allow_dispatity && (s.series == 0), // true, //false, // public boolean use_disparity; // adjust dx0+dx1+dy0+dy1 == 0
centerXY,
disparity_task,
disparity_meas,
disparity_task, // not used
disparity_meas, // not used
strength,
dxy)); // xy));
}
......@@ -975,13 +1096,15 @@ B = |+dy0 -dy1 -2*dy3 |
for (int i = 0; i< 8; i++) {
dbg_xy[i][nTile] /= dbg_xy[8][nTile];
}
dbg_xy[9][nTile] /= dbg_xy[8][nTile];
} else {
for (int i = 0; i< 8; i++) {
dbg_xy[i][nTile] = Double.NaN;
dbg_xy[9][nTile] = Double.NaN;
}
}
}
String [] titles = {"x0", "y0", "x1", "y1", "x2", "y2", "x3","y3","weight"};
String [] titles = {"x0", "y0", "x1", "y1", "x2", "y2", "x3","y3","weight","~disp"};
(new showDoubleFloatArrays()).showArrays(
dbg_xy,
tilesX,
......@@ -1341,6 +1464,7 @@ B = |+dy0 -dy1 -2*dy3 |
final double smplRms, // = 0.1; // Maximal RMS of the remaining tiles in a sample
final int tilesX)
{
final int dbg_tile = -34145; // 37005;
if (disp_strength_in.length > NUM_SLICES){
final double [][] disp_strength = new double [disp_strength_in.length][];
for (int nfirst = 0; nfirst < disp_strength_in.length; nfirst += NUM_SLICES){
......@@ -1377,6 +1501,9 @@ B = |+dy0 -dy1 -2*dy3 |
}
final double smlVar = smplRms * smplRms; // maximal variance (weighted average of the squared difference from the mean)
for (int nTile = 0; nTile < num_tiles; nTile++){
if (nTile == dbg_tile) {
System.out.println("filterDisparityStrength().1: nTile = dbg_tile = "+dbg_tile);
}
double w = disp_strength_in[1][nTile] - strength_floor;
if (w > 0){
if (strength_pow != 1.0) w = Math.pow(w, strength_pow);
......@@ -1385,6 +1512,10 @@ B = |+dy0 -dy1 -2*dy3 |
}
for (int tY = 0; tY < (tilesY - smplSide); tY++){
for (int tX = 0; tX < (tilesX - smplSide); tX++){
if ((tY*tilesX + tX + index_shift) == dbg_tile) {
System.out.println("filterDisparityStrength().2: nTile = dbg_tile = "+dbg_tile);
}
int num_in_sample = 0;
boolean [] smpl_sel = new boolean [smplLen];
double [] smpl_d = new double [smplLen];
......@@ -1394,7 +1525,7 @@ B = |+dy0 -dy1 -2*dy3 |
int y = tY + sy; // - smpl_center;
for (int sx = 0; sx < smplSide; sx++){
int x = tX + sx; // - smpl_center;
int indx = y * tilesX + x;
int indx = y * tilesX + x; // absolute center
if (weight[indx] > 0.0){
int indxs = sy * smplSide + sx;
smpl_sel[indxs] = true;
......@@ -1438,7 +1569,7 @@ B = |+dy0 -dy1 -2*dy3 |
}
}
if (iworst < 0){
System.out.println("**** this is a BUG in filterDisparityStrength() ****");
System.out.println("**** this is a BUG in filterDisparityStrength() ****"); // happened when smpl_d = [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]
break;
}
// remove worst sample
......@@ -1639,6 +1770,7 @@ B = |+dy0 -dy1 -2*dy3 |
public double [][][] lazyEyeCorrection(
final boolean use_poly, // Use polynomial correction, false - correct tilt/azimuth/roll of each sensor
final boolean restore_disp_inf, // Restore subtracted disparity for scan #0 (infinity)
final double fcorr_radius,
final double min_strength_in,
final double max_diff,
......@@ -1676,19 +1808,55 @@ B = |+dy0 -dy1 -2*dy3 |
// final double lazyEyeDispRelVariation = 0.02;
final int dbg_nTile = -59038;
final int dbg_nTile = -34145; // 37005; // -59038;
final int num_scans = scans_14.length/NUM_ALL_SLICES;
final int num_tiles = scans_14[0].length;
final int tilesY = num_tiles/tilesX;
final boolean [] center_mask = getCenterMask(fcorr_radius, tilesX, tilesY);
final double [][] scans = new double [num_scans * NUM_SLICES][];
final int [] indices_14_10 = {0,1,2,3,5,6,8,9,11,12};
final double [][] comp_strength_rms = new double [num_scans][num_tiles];
for (int ns = 0; ns < num_scans; ns++){
for (int i = 0; i < indices_14_10.length; i++){
scans[ns * NUM_SLICES + i] = scans_14[ns * NUM_ALL_SLICES + indices_14_10[i]];
final double [] min_weights = new double [num_tiles];
for (int nTile = 0; nTile < num_tiles; nTile++){
if (nTile == dbg_nTile) {
System.out.println("lazyEyeCorrection(), nTile="+nTile);
}
double w = scans_14[ns * NUM_ALL_SLICES + INDEX_14_WEIGHT][nTile];
for (int i = 0; i < INDICES_14_WEIGHTS.length; i++) {
w = Math.min(w, scans_14[ns * NUM_ALL_SLICES + INDICES_14_WEIGHTS[i]][nTile]);
}
min_weights[nTile] = w;
}
for (int i = 0; i < INDICES_14_10.length; i++){
if (i == INDEX_10_WEIGHT) {
scans[ns * NUM_SLICES + i] = min_weights;
} else {
scans[ns * NUM_SLICES + i] = scans_14[ns * NUM_ALL_SLICES + INDICES_14_10[i]];
}
}
}
if (debugLevel > -2) { // -2) { // 100) {
if (debugLevel > -3) { // -1) { // -2) { // 100) {
(new showDoubleFloatArrays()).showArrays(scans, tilesX, tilesY, true, "scans_pre-disp");
}
}
// Add disparity to dx0, dx1, dy2, dy3 pairs
if (restore_disp_inf) {
for (int nTile = 0; nTile < num_tiles; nTile++) if (scans[INDEX_10_WEIGHT][nTile] > 0){
for (int i = 0; i < INDICES_10_DISP.length; i++) {
scans[INDICES_10_DISP[i]][nTile] += scans[INDEX_10_DISPARITY][nTile];
}
}
}
if (debugLevel > -2) { // -2) { // 100) {
if (debugLevel > -3) { // -1) { // -2) { // 100) {
(new showDoubleFloatArrays()).showArrays(scans, tilesX, tilesY, true, "scans_post-disp");
}
}
// static final int [] INDICES_10_DISP = {2,4,7,9}; // which indices need to add disparity to restore full offsets
//INDEX_10_WEIGHT
for (int ns = 0; ns < num_scans; ns++){
for (int nTile = 0; nTile < num_tiles; nTile++){
......@@ -1715,9 +1883,9 @@ B = |+dy0 -dy1 -2*dy3 |
(new showDoubleFloatArrays()).showArrays(comp_strength_rms, tilesX, tilesY, true, "comp_strength_rms" , titles);
}
// FIXME: Seems that disparity should be combined with dxy for BG scan before that
double[][] filtered_scans = filterDisparityStrength (
scans, // final double[][] disp_strength_in,
scans, // final double[][] disp_strength_in, // [1][37006] >0, [21][37006] = NaN
min_strength_in, // final double strength_floor,
strength_pow, // final double strength_pow,
lazyEyeSmplSide, // final int smplSide, // = 2; // Sample size (side of a square)
......@@ -1727,7 +1895,7 @@ B = |+dy0 -dy1 -2*dy3 |
if (debugLevel > -2) { // -2) { // 100) {
(new showDoubleFloatArrays()).showArrays(filtered_scans, tilesX, tilesY, true, "filtered_scans");
if (debugLevel > -1) { // -2) { // 100) {
if (debugLevel > -3) { // -1) { // -2) { // 100) {
(new showDoubleFloatArrays()).showArrays(scans, tilesX, tilesY, true, "scans");
(new showDoubleFloatArrays()).showArrays(target_disparity, tilesX, tilesY, true, "target_disparity");
}
......@@ -1742,7 +1910,7 @@ B = |+dy0 -dy1 -2*dy3 |
for (int ns = 0; ns < num_scans; ns++){
for (int nTile = 0; nTile < num_tiles; nTile++) {
if (nTile == dbg_nTile){
System.out.println("lazyEyeCorrection().1: nTile="+nTile);
System.out.println("lazyEyeCorrection().1: nTile="+nTile); // filtered_scans[2][37005] = NaN
}
double w = filtered_scans[ns * NUM_SLICES + 1][nTile];
if (w > 0.0){
......@@ -1751,10 +1919,18 @@ B = |+dy0 -dy1 -2*dy3 |
for (int i = 2; i < NUM_SLICES; i++) if (i != 1){
combo_mismatch[i][nTile] += filtered_scans[ns * NUM_SLICES + i][nTile] * w;
}
//FIXME: ???? target_disparity is not 0 for bg
// combo_mismatch combines both infinity and regular for the same tile, mixing "disparity" and "target disparity" with weights and magic_scale
// Seems to be wrong, as target_disparity is only estimated disparity, not measured. Or is it measured for non-infinity?
// At least bg scan is measured with disparity =0, even as target_disparity is not 0
// combo data is later used as a non-infinity to correct all but disparity
if (target_disparity != null){
combo_mismatch[0][nTile] += (
filtered_scans[ns * NUM_SLICES + 0][nTile]/clt_parameters.corr_magic_scale +
target_disparity[ns][nTile])* w;
// target_disparity[ns][nTile])* w;
// disabling for bg - already combined
((ns==0)?0.0:target_disparity[ns][nTile])* w);
} else {
combo_mismatch[0][nTile] += (
filtered_scans[ns * NUM_SLICES + 0][nTile]/clt_parameters.corr_magic_scale +
......@@ -1786,21 +1962,20 @@ B = |+dy0 -dy1 -2*dy3 |
}
}
// reduce influence of high disparity
// reduce influence of high disparity, using combined disaprity
// double norm_ly_disparity = 100.0; // disabling
for (int nTile = 0; nTile < num_tiles; nTile++) {
if (combo_mismatch[0][nTile] > ly_norm_disp) {
if ((combo_mismatch[0][nTile] > 0) && (combo_mismatch[0][nTile] > ly_norm_disp)) {
combo_mismatch[1][nTile] *= ly_norm_disp/combo_mismatch[0][nTile];
}
}
if (debugLevel > 0) { // -2) { // 100) {
(new showDoubleFloatArrays()).showArrays(combo_comp_rms, tilesX, tilesY, "combo_comp_rms");
}
// instance of class to operate navigation over tiles
// compare tile disparity (combo) with those of neighbors, discard if too different
final TileNeibs tnImage = new TileNeibs(tilesX, tilesY); // num_tiles/tilesX);
for (int nTile = 0; nTile < num_tiles; nTile++) if (combo_mismatch[1][nTile] > 0.0){
if (nTile == dbg_nTile){
......@@ -1813,7 +1988,7 @@ B = |+dy0 -dy1 -2*dy3 |
if ((nTile1 >= 0) && (combo_mismatch[1][nTile1] > 0.0)){
if (Math.abs(combo_mismatch[0][nTile1] - d) > lev) { // azyEyeDispVariation){
combo_mismatch[1][nTile] = 0.0;
// for (int i = 0; i < NUM_SLICES; i++) if (i != 1){
combo_mismatch[0][nTile] = Double.NaN;
for (int i = 2; i < NUM_SLICES; i++) if (i != 1){
combo_mismatch[i][nTile] = Double.NaN;
}
......@@ -1823,6 +1998,8 @@ B = |+dy0 -dy1 -2*dy3 |
}
}
// here combo_mismatch[2][37005] = Double.NaN,combo_mismatch[1][37005] != 0.0, combo_mismatch[0][37005] = 0.0
if (debugLevel > 0) { // 0) {
String [] prefixes = {"disparity", "strength", "dx0", "dy0", "dx1", "dy1", "dx2", "dy2", "dx3", "dy3"};
(new showDoubleFloatArrays()).showArrays(combo_mismatch, tilesX, combo_mismatch[0].length/tilesX, true, "combo_mismatch" , prefixes);
......@@ -1848,9 +2025,10 @@ B = |+dy0 -dy1 -2*dy3 |
double [][] inf_scan = new double [NUM_SLICES][];
for (int i = 0; i < NUM_SLICES; i++){
inf_scan[i] = scans[i];
inf_scan[i] = scans[i]; // just copy first half (infinity data - now it has deltas x/y with restored disparity (no magic_scale)
}
// Optionally filter infinity scan data
if (smplSide > 1){
inf_scan = filterDisparityStrength (
inf_scan,
......@@ -1879,6 +2057,8 @@ B = |+dy0 -dy1 -2*dy3 |
}
}
// Optionally filter infinity scan data with a histogram filter
if (hist_smpl_side > 0) { // 0 to bypass histogram filtering
inf_scan = filterHistogramFar (
inf_scan, // final double[][] disp_strength_in,
......@@ -1913,8 +2093,21 @@ B = |+dy0 -dy1 -2*dy3 |
inf_and_ly[i + NUM_SLICES] = combo_mismatch[i];
}
// make all zero strength tiles to have NaN values to use histrograms in ImageJ
for (int ns = 0; ns < 2; ns++) {
for (int nt = 0; nt < inf_and_ly[INDEX_10_WEIGHT + ns * NUM_SLICES].length; nt++ ) {
if (inf_and_ly[INDEX_10_WEIGHT + ns * NUM_SLICES][nt] == 0.0) {
for (int i = 0; i < NUM_SLICES; i++) if (i != INDEX_10_WEIGHT){
inf_and_ly[i + ns * NUM_SLICES][nt] = Double.NaN;
}
}
}
}
// static final int INDEX_10_WEIGHT = 1;
if (debugLevel > 0) {
if (debugLevel > -1) { // 0) {
String [] prefixes = {"disparity", "strength", "dx0", "dy0", "dx1", "dy1", "dx2", "dy2", "dx3", "dy3"};
String [] titles = new String [2 * NUM_SLICES];
for (int i = 0; i < NUM_SLICES; i++){
......@@ -1999,13 +2192,15 @@ B = |+dy0 -dy1 -2*dy3 |
total_weights[0] += s.weight;
}
for (int nTile = 0; nTile < num_tiles; nTile++) if (center_mask[nTile]){
for (int nTile = 0; nTile < num_tiles; nTile++) if (center_mask[nTile]){ // calculate total weight of non-infinity
total_weights[1]+= inf_and_ly[1 * NUM_SLICES + 1][nTile];
}
double inf_fraction_limited = (inf_fraction >= 0.0) ?((inf_fraction > 1.0) ? 1.0 : inf_fraction):0.0;
double [] weights = {
inf_fraction * (total_weights[0] + total_weights[1]) / total_weights[0],
(1.0 - inf_fraction) * (total_weights[0] + total_weights[1]) / total_weights[1],
inf_fraction_limited * (total_weights[0] + total_weights[1]) / total_weights[0],
(1.0 - inf_fraction_limited) * (total_weights[0] + total_weights[1]) / total_weights[1],
};
for (int ns = 0; ns <2; ns++) {
......@@ -2051,18 +2246,31 @@ B = |+dy0 -dy1 -2*dy3 |
}
if (debugLevel > 0) {
if (debugLevel > -1) {
String [] prefixes = {"disparity", "strength", "dx0", "dy0", "dx1", "dy1", "dx2", "dy2", "dx3", "dy3"};
(new showDoubleFloatArrays()).showArrays(combo_mismatch, tilesX, combo_mismatch[0].length/tilesX, true, "combo_mismatch" , prefixes);
}
if (debugLevel > -1) { // 0) {
String [] prefixes = {"disparity", "strength", "dx0", "dy0", "dx1", "dy1", "dx2", "dy2", "dx3", "dy3"};
String [] titles = new String [2 * NUM_SLICES];
for (int i = 0; i < NUM_SLICES; i++){
titles[i] = prefixes[i]+"-inf";
titles[i + NUM_SLICES] = prefixes[i]+"-ly";
}
(new showDoubleFloatArrays()).showArrays(inf_and_ly, tilesX, tilesY, true, "inf_and_ly_last",titles);
}
ArrayList<Mismatch> mismatch_list = use_poly? null : (new ArrayList<Mismatch>());
// inf_and_ly here has filtered disparity and offsets, should be process clt_parameters.ly_inf_disp before filters
double [][][] mismatch_corr_coefficients = infinityMismatchCorrection(
clt_parameters.disp_scan_start, // final double disp_scan_start,
clt_parameters.disp_scan_step, // final double disp_scan_step,
clt_parameters.disp_scan_step, // final double disp_scan_step,
use_poly, // final boolean use_poly,
clt_parameters.fcorr_quadratic, // final boolean use_quadratic,
true, // clt_parameters.fcorr_inf_vert, // final boolean use_vertical,
// tool alte to restore disparity - should be dome earlier
false, // final boolean use_disparity, // for infinity
clt_parameters.ly_inf_disp, //final boolean allow_dispatity,
clt_parameters, // EyesisCorrectionParameters.CLTParameters clt_parameters,
inf_and_ly, // double [][] disp_strength,
inf_samples_list, // ArrayList<Sample> samples_list,
......@@ -2076,6 +2284,8 @@ B = |+dy0 -dy1 -2*dy3 |
show_fine_corr(
mismatch_corr_coefficients,
"mismatch_corr_coefficients");
} else {
System.out.println("Are null - non-null are for poly correction only");
}
}
if (!use_poly && (mismatch_list != null)){
......@@ -2201,6 +2411,15 @@ B = |+dy0 -dy1 -2*dy3 |
}
sdfa_instance.showArrays(dbg_clt_mismatch, tilesX, tilesY, true, title, titles);
}
/**
* Calculates transposed Jacobian for 8*<number of tiles> points (port x,y coordinates) for each of the symmetrical parameters (starting with zoom)
* @param par_mask bitmask of selected parameters, starting with sym0 (convergence)
* @param mismatch_list list of samples
* @param geometryCorrection instance of the camera geometry class
* @param corr_vector current correction vector
* @param debugLevel debug level
* @return transposed Jacobian
*/
double [][] getJacobianTransposed(
boolean [] par_mask,
......@@ -2251,12 +2470,12 @@ B = |+dy0 -dy1 -2*dy3 |
mm.getDisparityMeas()); // getDisparityTask()); // double disparity)
// convert to symmetrical coordinates
// derivatives of each port coordinates (in pixels) for each of selected symmetric all parameters (sym0 is convergence for disparity)
double [][] jt_partial = corr_vector.getJtPartial(
deriv, // double [][] port_coord_deriv,
par_mask); // boolean [] par_mask
// put partial transposed jacobian into full transposed Jacobian
// put partial (for 1 tile - 8 port coordinates) transposed jacobian into full (all tiles) transposed Jacobian
for (int npar = 0; npar < jt.length; npar++){
for (int n = 0; n < 2* NUM_SENSORS; n++){
// jt[npar][2 * NUM_SENSORS * indx + n] = j_partial[n][npar]; // here Jacobian was not transposed
......@@ -2526,6 +2745,11 @@ B = |+dy0 -dy1 -2*dy3 |
double rms = 0.0;
double sw = 0.0;
for (int i = 0; i < w.length; i++){
if (Double.isNaN(y[i])) {
// TODO: Fix source !
System.out.println("getRMS(): y["+i+"]="+y[i]+", w[i]="+w[i]);
w[i] = 0.0;
}
rms += w[i]*y[i]*y[i];
sw += w[i];
}
......@@ -2535,13 +2759,36 @@ B = |+dy0 -dy1 -2*dy3 |
return rms;
}
/**
* Create Y-F(x) vector for 8 points per tile data from the mismatch_list. Groups of 8 uses linear combinations of the measured {dx0, dy0,..dx3,dy3},
* and the first 7 points in each group are invariant of disparity, the 8-th is optionally disabled by the corresponding weight (for points with unknown disparity)
*
* Actually there is no explicit F(x), because after each step, the correlations are recalculated instead, and the measured are already differences from "F(x)"
*
* @param mismatch_list list of the tile samples
* @return array of 8*<number of tiles> values
*/
double [] getYminusFx(
ArrayList<Mismatch> mismatch_list)
{
double [] yMinusFx = new double [2 * NUM_SENSORS * mismatch_list.size()];
double [] w_DBG = new double [2 * NUM_SENSORS * mismatch_list.size()];
for (int indx = 0; indx<mismatch_list.size(); indx++){ // need indx value
Mismatch mm = mismatch_list.get(indx);
mm.copyToY(yMinusFx, indx);
mm.copyToW(w_DBG, indx); // temporarily
}
// check for NaN:
// temporary
for (int i = 0; i < yMinusFx.length; i++) {
if (Double.isNaN(yMinusFx[i]) && (w_DBG[i] != 0 )) {
System.out.println("**** getYminusFx BUG! y["+i+"] = NaN, while w is != 0 *****");
System.out.println("**** getYminusFx BUG! y["+i+"] = NaN, while w is != 0 *****");
}
}
return yMinusFx;
}
......@@ -2550,7 +2797,7 @@ B = |+dy0 -dy1 -2*dy3 |
ArrayList<Mismatch> mismatch_list)
{
double [] w = new double [2 * NUM_SENSORS * mismatch_list.size()];
for (int indx = 0; indx<mismatch_list.size(); indx++){ // need indx value
for (int indx = 0; indx < mismatch_list.size(); indx++){ // need indx value
Mismatch mm = mismatch_list.get(indx);
mm.copyToW(w, indx);
}
......@@ -2587,11 +2834,13 @@ B = |+dy0 -dy1 -2*dy3 |
}
boolean [] par_mask = geometryCorrection.getParMask(
has_disparity, // boolean use_disparity,
// temporary - just for testing
force_convergence, // boolean disparity_only,
force_convergence && has_disparity, // boolean use_disparity,
common_roll,// boolean common_roll,
corr_focalLength); // boolean corr_focalLength);
double [][] jta = getJacobianTransposed(
double [][] jta = getJacobianTransposed( // gets transposed jacobian for 8*num_tiles points (port coordinates) and symmetrical parameters
par_mask, // boolean [] par_mask,
mismatch_list, // ArrayList<Mismatch> mismatch_list,
geometryCorrection, // GeometryCorrection geometryCorrection,
......@@ -2600,15 +2849,15 @@ B = |+dy0 -dy1 -2*dy3 |
// debugLevel = 2;
// convert Jacobian outputs to symmetrical measurement vectors (last one is non-zero only if disparity should be adjusted)
// Convert Jacobian outputs to symmetrical measurement vectors (last one is non-zero only if disparity should be adjusted)
// so now each group of 8 points (individual coordinate pairs for each sub-camera) is replaced by 8-point "measurement vector", where only the last component
// depends on disparity. That last component is optionally disabled by the corresponding weight for the samples that do not have absolute disparity data
double [][] jta_mv = (new Mismatch()).convertJt_mv (jta); //double [][] jt)
Matrix jt = new Matrix(jta_mv);
Matrix jt = new Matrix(jta_mv);
// Extract measured offsets differences from the list and convert them to the measured vectors for which the jta is intended
double [] y_minus_fx_a = getYminusFx( // mv[0]..mv[7], not the measured data (dx0, dy0, ... dx3, dy3)
mismatch_list); // ArrayList<Mismatch> mismatch_list)
......@@ -2616,7 +2865,7 @@ B = |+dy0 -dy1 -2*dy3 |
mismatch_list); // ArrayList<Mismatch> mismatch_list)
double [] y_minus_fx_a_weighted = mulWeight(y_minus_fx_a, weights);
double rms0 = getRMS (y_minus_fx_a, weights);
if (debugLevel > -2){
if (debugLevel > -3){
System.out.println("--- solveCorr(): initial RMS = " + rms0);
}
......@@ -2635,7 +2884,7 @@ B = |+dy0 -dy1 -2*dy3 |
int dbg_length = dbg_owidth*dbg_oheight;
String [] dbg_titles_sym= {"sym0","sym1","sym2","sym3","sym4","sym5","sroll0","sroll1","sroll2","sroll3", "zoom0", "zoom1", "zoom2"};
String [] dbg_titles_xy= {"x0","y0","x1","y1","x2","y2","x3","y3"};
String [] dbg_titles_mv= {"dy0","dy1","dx2","dx3","dx1-dx0","dy3-dy2","dh-dv","dhy+dv"};
String [] dbg_titles_mv= {"dy0","dy1","dx2","dx3","dx1-dx0","dy3-dy2","dh-dv","dh+dv"};
double [][] dbg_xy = null; // jacobian dmv/dsym
double [][] dbg_mv = null; // jacobian dmv/dsym
double [][] dbg_dmv_dsym = null; // jacobian dmv/dsym
......@@ -2694,7 +2943,7 @@ B = |+dy0 -dy1 -2*dy3 |
}
dbgImgRemoveEmpty(dbg_xy);
dbgImgRemoveEmpty(dbg_xy);
dbgImgRemoveEmpty(dbg_mv);
dbgImgRemoveEmpty(dbg_dmv_dsym);
dbgImgRemoveEmpty(dbg_dmv_dsym_delta);
dbgImgRemoveEmpty(dbg_dmv_dsym_diff);
......@@ -2717,13 +2966,14 @@ B = |+dy0 -dy1 -2*dy3 |
for (int i = 0; i < drslt.length; i++){
drslt[i] *= -1.0;
}
//if (par_mask[0]) drslt[0] *= -1.0; //FIXME: Find actual bug, sym[0] corrects in opposite way
GeometryCorrection.CorrVector rslt = geometryCorrection.getCorrVector(drslt, par_mask);
if (debugLevel > -2){
if (debugLevel > -3){ // change to >0) {
System.out.println("solveCorr() rslt:");
System.out.println(rslt.toString());
}
return rslt;
}
......
src/main/java/Correlation2d.java
View file @ b6994b6e
......@@ -704,7 +704,7 @@ public class Correlation2d {
if (Double.isNaN(data[imx]) || (data[i] > data[imx])) imx = i;
}
}
if (data[imx] < minMax) {
if (!(data[imx] >= minMax)) {
if (debug){
System.out.println("getMaxXYInt() -> null (data["+imx+"] = "+data[imx]+" < "+minMax);
}
......@@ -963,6 +963,101 @@ public class Correlation2d {
return padded_strip;
}
public double [] mismatchPairsCM( // returns x-xcenter, y, strength (sign same as disparity)
ImageDttParameters imgdtt_params,
double [][] corrs,
int pair_mask, // which pairs to process
double xcenter, // -disparity to compare
double radius, // positive - within that distance, negative - within 2*(-radius)+1 square
int debug_level,
int tileX, // just for debug output
int tileY)
{
boolean debug = debug_level > 0;
int width = 2 * transform_size - 1;
int center = transform_size - 1;
int center_index = (width + 1) * center; //
int num_pairs = 0;
for (int i = 0; i < corrs.length; i++) if ((corrs[i] != null) && (((1 << i) & pair_mask) != 0)) {
num_pairs++;
}
double [] rslt = new double[3 * num_pairs];
int np= 0;
int [] icenter = new int[2];
int ixcenter = (int) Math.round(xcenter);
for (int pair = 0; pair < corrs.length; pair++) if ((corrs[pair] != null) && (((1 << pair) & pair_mask) != 0)) {
// int this_mask = 1 << pair;
if ( isHorizontalPair(pair)) {
icenter[0] = ixcenter;
icenter[1] = 0;
} else if (isVerticalPair(pair)) {
icenter[0] = 0;
icenter[1] = ixcenter;
} else if (isDiagonalMainPair(pair)) {
icenter[0] = ixcenter;
icenter[1] = ixcenter;
} else if (isDiagonalOtherPair(pair)) {
icenter[0] = ixcenter;
icenter[1] = -ixcenter;
} else {
System.out.println("************ BUG: illegal pair type for pair1"+pair);
return null;
}
//calculate as "center of mass"
int iradius = (int) Math.abs(radius);
int ir2 = (int) (radius*radius);
boolean square = radius <0;
double s0 = 0, sx=0, sy = 0;
for (int y = - iradius ; y <= iradius; y++){
int dataY = icenter[1] +y;
if ((dataY >= -center) && (dataY <= center)){
int y2 = y*y;
for (int x = - iradius ; x <= iradius; x++){
int dataX = icenter[0] +x;
double r2 = y2 + x * x;
if ((dataX >= -center) && (dataX <= center) && (square || (r2 <= ir2))){
double d = corrs[pair][dataY * width + dataX + center_index];
if (d > 0.0) {
s0 += d;
sx += d * dataX;
sy += d * dataY;
}
}
}
}
}
double xm = sx / s0;
double ym = sy / s0;
int ixm = (int) Math.round(xm);
int iym = (int) Math.round(ym);
double s = corrs[pair][iym * width + ixm + center_index];
if ( isHorizontalPair(pair)) {
rslt[3 * np + 0] = xcenter - xm;
rslt[3 * np + 1] = -ym;
} else if (isVerticalPair(pair)) {
rslt[3 * np + 0] = -xm;
rslt[3 * np + 1] = xcenter - ym;
} else if (isDiagonalMainPair(pair)) {
rslt[3 * np + 0] = xcenter - xm;
rslt[3 * np + 1] = xcenter - ym;
} else if (isDiagonalOtherPair(pair)) {
rslt[3 * np + 0] = xcenter - xm;
rslt[3 * np + 1] = -xcenter - ym;
} else {
System.out.println("************ BUG: illegal pair type for pair "+pair);
return null;
}
rslt[3 * np + 2] = s;
if (debug){
System.out.println("getMaxXYInt() -> "+rslt[0]+"/"+rslt[1]);
}
np++;
}
return rslt;
}
// returns array 3*num_pairs long
// TODO: now works for small offsets. Maybe add re-calculate int argmax for each pair? xcenter is still needed to subtract Add switch? (small/large correction)
......@@ -983,9 +1078,9 @@ public class Correlation2d {
double [] rslt = new double[3 * num_pairs];
int np= 0;
for (int i = 0; i < corrs.length; i++) if ((corrs[i] != null) && (((1 << i) & pair_mask) != 0)) {
for (int pair = 0; pair < corrs.length; pair++) if ((corrs[pair] != null) && (((1 << pair) & pair_mask) != 0)) {
// for (int np = 0; np < num_pairs; np++) {
int this_mask = 1 << i;
int this_mask = 1 << pair;
Correlations2dLMA lma=corrLMA(
imgdtt_params, // ImageDttParameters imgdtt_params,
corrs, // double [][] corrs,
......@@ -1000,23 +1095,43 @@ public class Correlation2d {
if ((lma == null) || (lma.getPoly() == null)) {
rslt[3 * np + 0] = Double.NaN;
rslt[3 * np + 1] = Double.NaN;
rslt[3 * np + 2] = Double.NaN;
rslt[3 * np + 2] = 0.0;
} else {
double [] poly_xyvwh = lma.getPoly();
rslt[3 * np + 0] = xcenter - poly_xyvwh[0];
rslt[3 * np + 1] = -poly_xyvwh[1];
rslt[3 * np + 2] = poly_xyvwh[2];
if ( isHorizontalPair(pair)) {
rslt[3 * np + 0] = xcenter - poly_xyvwh[0];
rslt[3 * np + 1] = -poly_xyvwh[1];
} else if (isVerticalPair(pair)) {
rslt[3 * np + 0] = -poly_xyvwh[1];
rslt[3 * np + 1] = xcenter - poly_xyvwh[0];
} else if (isDiagonalMainPair(pair)) {
// rslt[3 * np + 0] = xcenter - poly_xyvwh[0];
// rslt[3 * np + 1] = -poly_xyvwh[1];
rslt[3 * np + 0] = xcenter - poly_xyvwh[0] + poly_xyvwh[1]; // x - y
rslt[3 * np + 1] = xcenter - poly_xyvwh[0] - poly_xyvwh[1]; // x + y
} else if (isDiagonalOtherPair(pair)) {
// rslt[3 * np + 0] = xcenter - poly_xyvwh[0];
// rslt[3 * np + 1] = poly_xyvwh[1];
rslt[3 * np + 0] = xcenter - poly_xyvwh[0] + poly_xyvwh[1]; // x - y
rslt[3 * np + 1] = -xcenter + poly_xyvwh[0] + poly_xyvwh[1]; // x + y
} else {
System.out.println("************ BUG: illegal pair type for pair "+pair);
return null;
}
rslt[3 * np + 2] = Double.isNaN(poly_xyvwh[2])?0.0: poly_xyvwh[2];
}
if (debug_level > 0) {
System.out.println(String.format("mismatchPairs(), np = %d pairs mask = 0x%x, dx=%f, dy=%f, strength=%f", np, this_mask, rslt[3 * np + 0], rslt[3 * np + 1], rslt[3 * np + 2]));
if ((Double.isNaN(rslt[3 * np + 0]) || Double.isNaN(rslt[3 * np + 1])) && (rslt[3 * np + 2] > 0.0)) {
System.out.println("************ mismatchPairs() Fix NaN!!!!! **************");
System.out.println(String.format("(), np = %d pairs mask = 0x%x, dx=%f, dy=%f, strength=%f", np, this_mask, rslt[3 * np + 0], rslt[3 * np + 1], rslt[3 * np + 2]));
}
if (debug_level > -1) {
System.out.println(String.format("(), np = %d pairs mask = 0x%x, dx=%f, dy=%f, strength=%f", np, this_mask, rslt[3 * np + 0], rslt[3 * np + 1], rslt[3 * np + 2]));
}
np++;
}
return rslt;
}
public double [][] corr4dirsLMA(
ImageDttParameters imgdtt_params,
double [][] corrs,
......
src/main/java/EyesisCorrectionParameters.java
View file @ b6994b6e
......@@ -2459,6 +2459,7 @@ public class EyesisCorrectionParameters {
public double ly_inf_frac = 0.5; // Relative weight of infinity calibration data
public boolean ly_on_scan = true; // Calculate and apply lazy eye correction after disparity scan (poly or extrinsic)
public boolean ly_inf_en = false; // true; // Simultaneously correct disparity at infinity (both poly and extrinsic)
public boolean ly_inf_disp= false; // Correct disparity for infinity tiles
public boolean ly_inf_force= false; // Force convergence correction during extrinsic, even with no infinity data
public boolean ly_com_roll= false; // Enable common roll (valid for high disparity range only)
......@@ -3140,6 +3141,8 @@ public class EyesisCorrectionParameters {
properties.setProperty(prefix+"ly_inf_frac", this.ly_inf_frac +"");
properties.setProperty(prefix+"ly_on_scan", this.ly_on_scan+"");
properties.setProperty(prefix+"ly_inf_en", this.ly_inf_en+"");
properties.setProperty(prefix+"ly_inf_disp", this.ly_inf_disp+"");
properties.setProperty(prefix+"ly_inf_force", this.ly_inf_force+"");
properties.setProperty(prefix+"ly_com_roll", this.ly_com_roll+"");
properties.setProperty(prefix+"ly_focalLength", this.ly_focalLength+"");
......@@ -3770,6 +3773,7 @@ public class EyesisCorrectionParameters {
if (properties.getProperty(prefix+"ly_inf_frac")!=null) this.ly_inf_frac=Double.parseDouble(properties.getProperty(prefix+"ly_inf_frac"));
if (properties.getProperty(prefix+"ly_on_scan")!=null) this.ly_on_scan=Boolean.parseBoolean(properties.getProperty(prefix+"ly_on_scan"));
if (properties.getProperty(prefix+"ly_inf_en")!=null) this.ly_inf_en=Boolean.parseBoolean(properties.getProperty(prefix+"ly_inf_en"));
if (properties.getProperty(prefix+"ly_inf_disp")!=null) this.ly_inf_disp=Boolean.parseBoolean(properties.getProperty(prefix+"ly_inf_disp"));
if (properties.getProperty(prefix+"ly_inf_force")!=null) this.ly_inf_force=Boolean.parseBoolean(properties.getProperty(prefix+"ly_inf_force"));
if (properties.getProperty(prefix+"ly_com_roll")!=null) this.ly_com_roll=Boolean.parseBoolean(properties.getProperty(prefix+"ly_com_roll"));
if (properties.getProperty(prefix+"ly_focalLength")!=null) this.ly_focalLength=Boolean.parseBoolean(properties.getProperty(prefix+"ly_focalLength"));
......@@ -4431,12 +4435,15 @@ public class EyesisCorrectionParameters {
gd.addMessage ("Maximal measured relative disparity = "+ (0.8*disp_scan_step)+" (0.8 * disp_scan_step)");
// gd.addNumericField("Maximal measured relative disparity", this.ly_meas_disp, 3);
gd.addNumericField("Maximal RMS of the remaining tiles in a sample", this.ly_smpl_rms, 5);
gd.addNumericField("Maximal full disparity difference to 8 neighbors", this.ly_disp_var, 5);
gd.addNumericField("Maximal relative full disparity difference to 8 neighbors", this.ly_disp_rvar, 5);
gd.addNumericField("Maximal full disparity difference to 8 neighbors", this.ly_disp_var, 8,5,"pix",
"Full allowed mismatch is a sum of absolute and disparity times relative");
gd.addNumericField("Maximal relative full disparity difference to 8 neighbors", this.ly_disp_rvar, 8,5,"",
"Full allowed mismatch is a sum of absolute and disparity times relative");
gd.addNumericField("Reduce weight of higher disparity tiles", this.ly_norm_disp, 5);
gd.addNumericField("Relative weight of infinity calibration data", this.ly_inf_frac, 3);
gd.addCheckbox ("Calculate and apply lazy eye correction after disparity scan (poly or extrinsic), may repeat",this.ly_on_scan);
gd.addCheckbox ("Use infinity disparity (disable if there is not enough of infinity data), both poly and extrinsic", this.ly_inf_en);
gd.addCheckbox ("Correct disparity for infinity tiles )has to disable until code fixed)", this.ly_inf_disp);
gd.addCheckbox ("Force convergence correction during extrinsic, even with no infinity data", this.ly_inf_force);
gd.addCheckbox ("Enable common roll adjustment (valid for high disparity range scans only)", this.ly_com_roll);
gd.addCheckbox ("Correct scales (focal length temperature? variations)", this.ly_focalLength);
......@@ -5139,6 +5146,7 @@ public class EyesisCorrectionParameters {
this.ly_inf_frac= gd.getNextNumber();
this.ly_on_scan= gd.getNextBoolean();
this.ly_inf_en= gd.getNextBoolean();
this.ly_inf_disp= gd.getNextBoolean();
this.ly_inf_force= gd.getNextBoolean();
this.ly_com_roll= gd.getNextBoolean();
this.ly_focalLength= gd.getNextBoolean();
......@@ -5790,12 +5798,12 @@ public class EyesisCorrectionParameters {
if (z_corr_map.size() > 0) {
gd.addMessage("Edit infinity disparity correction (in 1/m), set >= 1.0 to remove for the following");
for (HashMap.Entry<String,Double> entry : z_corr_map.entrySet()){
gd.addNumericField(entry.getKey(), entry.getValue(), 8);
gd.addNumericField(entry.getKey(), entry.getValue(), 9,12, "m-1");
}
}
gd.addMessage("Add new infinity correction");
gd.addStringField ("Timestamp string (seconds_microseconds):", "", 40);
gd.addNumericField("Infinity correction (in 1/m)", 0, 8);
gd.addNumericField("Infinity correction (in 1/m)", 0, 9,12,"m-1");
gd.addCheckbox ("Clear list", false);
WindowTools.addScrollBars(gd);
gd.showDialog();
......
src/main/java/GeometryCorrection.java
View file @ b6994b6e
......@@ -101,11 +101,13 @@ public class GeometryCorrection {
}
public boolean [] getParMask(
boolean disparity_only,
boolean use_disparity,
boolean common_roll,
boolean corr_focalLength)
{
return (new CorrVector()).getParMask(
disparity_only,
use_disparity,
common_roll,
corr_focalLength);
......@@ -378,7 +380,7 @@ public class GeometryCorrection {
s += " |↘ ↙| |↘ ↗| |↗ ↘| |↙ ↘| |↙ ↗| |↖ ↘| 6: common roll 7:(r0-r3)/2, |- +| |- -| |- +|\n";
s += " 0: |↗ ↖| 1: |↙ ↖| 2: |↖ ↙| 3: |↖ ↗| 4: |↗ ↙| 5: |↘ ↖| 8:(r1-r2)/2 9:(r0+r3-r1-r2)/4 10: |- +| 11: |+ +| 12: |+ -|\n";
s += String.format(" 0: %8.5f° 1:%8.5f° 2:%8.5f° 3:%8.5f° 4:%8.5f° 5:%8.5f° 6:%8.5f° 7:%8.5f° 8:%8.5f° 9:%8.5f° 10: %8.5f‰ 11:%8.5f‰ 12:%8.5f‰\n" ,
s += String.format(" 0:%9.6f° 1:%8.5f° 2:%8.5f° 3:%8.5f° 4:%8.5f° 5:%8.5f° 6:%8.5f° 7:%8.5f° 8:%8.5f° 9:%8.5f° 10: %8.5f‰ 11:%8.5f‰ 12:%8.5f‰\n" ,
sv[0], sv[1], sv[2], sv[3], sv[4], sv[5], sv[6], sv[7], sv[8], sv[9], 1000*sv[10], 1000*sv[11], 1000*sv[12] );
return s;
......@@ -506,21 +508,25 @@ matrix([[-0.125, -0.125, 0.125, 0.125, -0.125, 0.125, -0. , -0. , -0.
}
public boolean [] getParMask(
boolean disparity_only,
boolean use_disparity,
boolean common_roll,
boolean corr_focalLength)
{
common_roll &= !disparity_only;
corr_focalLength &= !disparity_only;
use_disparity |= disparity_only;
boolean [] par_mask = {
use_disparity, //sym0
true, //sym1
true, //sym2
true, //sym3
true, //sym4
true, //sym5
!disparity_only, //sym1
!disparity_only, //sym2
!disparity_only, //sym3
!disparity_only, //sym4
!disparity_only, //sym5
common_roll, //sym6 // common roll
true, //sym7
true, //sym8
true, //sym9
!disparity_only, //sym7
!disparity_only, //sym8
!disparity_only, //sym9
corr_focalLength, //sym10
corr_focalLength, //sym11
corr_focalLength //sym12
......@@ -533,7 +539,7 @@ matrix([[-0.125, -0.125, 0.125, 0.125, -0.125, 0.125, -0. , -0. , -0.
* Get partial transposed Jacobian as 2d array (for one measurement set) from partial Jacobian for each sample
* with derivatives of port coordinates (all 4) by 3 tilts (ports 0..2), 3 azimuths (ports 0..2) and all 4 rolls
* Tilt and azimuth for port 3 is calculated so center would not move. Tilt is positive up, azimuth - right and
* roll - clockwise
* roll - clockwise. Added zooms (difference from 1.0) for sensors 0..2
*
* Result is transposed Jacobian (rows (9 , 10,12 or 13) - parameters, columns - port coordinate components (8). Parameters
* here are symmetrical, 0 is disparity-related (all to the center), remaining 9 preserve disparity and
......@@ -1225,7 +1231,7 @@ matrix([[-0.125, -0.125, 0.125, 0.125, -0.125, 0.125, -0. , -0. , -0.
double dpYci_droll = drvi_drl.get(1, 0) * norm_z - pYci * drvi_drl.get(2, 0) / rvi.get(2, 0);
double dpXci_dzoom = drvi_dzm.get(0, 0) * norm_z - pXci * drvi_dzm.get(2, 0) / rvi.get(2, 0);
double dpYci_dzoom = drvi_drl.get(1, 0) * norm_z - pYci * drvi_dzm.get(2, 0) / rvi.get(2, 0);
double dpYci_dzoom = drvi_dzm.get(1, 0) * norm_z - pYci * drvi_dzm.get(2, 0) / rvi.get(2, 0);
double dri_dazimuth = ri_scale / rNDi* (pXci * dpXci_dazimuth + pYci * dpYci_dazimuth);
double dri_dtilt = ri_scale / rNDi* (pXci * dpXci_dtilt + pYci * dpYci_dtilt);
......@@ -1258,22 +1264,22 @@ matrix([[-0.125, -0.125, 0.125, 0.125, -0.125, 0.125, -0. , -0. , -0.
// verify that d/dsym are well, symmetrical
if (i < (numSensors - 1)){
pXYderiv[2 * i + 0][CorrVector.TILT_INDEX+i] = dpXid_dtilt;
pXYderiv[2 * i + 1][CorrVector.TILT_INDEX+i] = dpYid_dtilt;
pXYderiv[2 * i + 0][CorrVector.AZIMUTH_INDEX+i] = dpXid_dazimuth;
pXYderiv[2 * i + 1][CorrVector.AZIMUTH_INDEX+i] = dpYid_dazimuth;
pXYderiv[2 * i + 0][CorrVector.TILT_INDEX+i] = dpXid_dtilt;
pXYderiv[2 * i + 1][CorrVector.TILT_INDEX+i] = dpYid_dtilt;
pXYderiv[2 * i + 0][CorrVector.AZIMUTH_INDEX+i] = dpXid_dazimuth;
pXYderiv[2 * i + 1][CorrVector.AZIMUTH_INDEX+i] = dpYid_dazimuth;
pXYderiv[2 * i + 0][CorrVector.ZOOM_INDEX+i] = dpXid_dzoom;
pXYderiv[2 * i + 1][CorrVector.ZOOM_INDEX+i] = dpYid_dzoom;
pXYderiv[2 * i + 0][CorrVector.ZOOM_INDEX+i] = dpXid_dzoom;
pXYderiv[2 * i + 1][CorrVector.ZOOM_INDEX+i] = dpYid_dzoom;
} else {
for (int j = 0; j < (numSensors - 1); j++){
pXYderiv[2 * i + 0][CorrVector.TILT_INDEX+j] = -dpXid_dtilt;
pXYderiv[2 * i + 1][CorrVector.TILT_INDEX+j] = -dpYid_dtilt;
pXYderiv[2 * i + 0][CorrVector.TILT_INDEX+j] = -dpXid_dtilt;
pXYderiv[2 * i + 1][CorrVector.TILT_INDEX+j] = -dpYid_dtilt;
pXYderiv[2 * i + 0][CorrVector.AZIMUTH_INDEX+j] = -dpXid_dazimuth;
pXYderiv[2 * i + 1][CorrVector.AZIMUTH_INDEX+j] = -dpYid_dazimuth;
pXYderiv[2 * i + 0][CorrVector.ZOOM_INDEX+j] = -dpXid_dzoom;
pXYderiv[2 * i + 1][CorrVector.ZOOM_INDEX+j] = -dpYid_dzoom;
pXYderiv[2 * i + 0][CorrVector.ZOOM_INDEX+j] = -dpXid_dzoom;
pXYderiv[2 * i + 1][CorrVector.ZOOM_INDEX+j] = -dpYid_dzoom;
}
}
pXYderiv[2 * i + 0][CorrVector.ROLL_INDEX+i] = dpXid_droll;
......
src/main/java/ImageDtt.java
View file @ b6994b6e
......@@ -1493,7 +1493,7 @@ public class ImageDtt {
if (i == (transform_size-1)) enh_ortho_scale[i] = 0.0 ; // hardwired 0 in the center
enh_ortho_scale[i] *= Math.sin(Math.PI*(i+1.0)/(2*transform_size));
}
if (globalDebugLevel > 0){
if (globalDebugLevel > 1){
System.out.println("enhortho_width="+ imgdtt_params.enhortho_width+" enhortho_scale="+ imgdtt_params.enhortho_scale);
for (int i = 0; i < corr_size; i++){
System.out.println(" enh_ortho_scale["+i+"]="+ enh_ortho_scale[i]);
......@@ -1927,7 +1927,7 @@ public class ImageDtt {
}
}
int tile_lma_debug_level = ((tileX == debug_tileX) && (tileY == debug_tileY))?imgdtt_params.lma_debug_level : -1;
int tile_lma_debug_level = ((tileX == debug_tileX) && (tileY == debug_tileY))? imgdtt_params.lma_debug_level : -1;
// all color channels are done here
double extra_disparity = 0.0; // used for textures: if allowed, shift images extra before trying to combine
......@@ -1935,7 +1935,10 @@ public class ImageDtt {
// fill clt_corr_combo if it exists
if (disparity_map != null){ // not null - calculate correlations
for (int i = 0; i < disparity_map.length; i++) {
if (disparity_map[i] != null) disparity_map[i][nTile] = Double.NaN; // once and for all
if (disparity_map[i] != null) disparity_map[i][nTile] = (
(i == DISPARITY_STRENGTH_INDEX) ||
(i == DISPARITY_INDEX_HOR_STRENGTH) ||
(i == DISPARITY_INDEX_VERT_STRENGTH)) ? 0.0 : Double.NaN; // once and for all
}
//clt_mismatch should only be used with disparity_map != null;
if (clt_mismatch != null) {
......@@ -2056,6 +2059,9 @@ public class ImageDtt {
disparity_map[DISPARITY_INDEX_INT][tIndex] = -ixy[0];
// disparity_map[DISPARITY_INDEX_INT + 1][tIndex] =
disparity_map[DISPARITY_STRENGTH_INDEX][tIndex] = strength;
if (Double.isNaN(disparity_map[DISPARITY_STRENGTH_INDEX][tIndex])) {
System.out.println("BUG: 1. disparity_map[DISPARITY_STRENGTH_INDEX]["+tIndex+"] should not be NaN");
}
corr_stat = corr2d.getMaxXCm( // get fractional center as a "center of mass" inside circle/square from the integer max
strip_combo, // double [] data, // [data_size * data_size]
ixy[0], // int ixcenter, // integer center x
......@@ -2140,6 +2146,9 @@ public class ImageDtt {
strength = lma_disparity_strength[1];
disparity_map[DISPARITY_INDEX_CM] [tIndex] = disparity;
disparity_map[DISPARITY_STRENGTH_INDEX] [tIndex] = strength;
if (Double.isNaN(disparity_map[DISPARITY_STRENGTH_INDEX][tIndex])) {
System.out.println("BUG: 2. disparity_map[DISPARITY_STRENGTH_INDEX][tIndex] should not be NaN");
}
}
// store debug data
// if strong enough and enabled - try to improve far objects
......@@ -2203,20 +2212,38 @@ public class ImageDtt {
}
if (tile_lma_debug_level > -1) {
System.out.println("debug12348973591");
}
if (clt_mismatch != null) { // mod_disparity_diff should be calculated
// bypass difference or zero strength if disparity difference is too high (will influence mismatch correction)
// but setting it too low will make it impossible to correct larger mismatches. Maybe multi-pass?
if (mod_disparity_diff[2] <= imgdtt_params.mismatch_max_diff) { // may be NaN, will fail test as intended
double [] mismatch_result = corr2d.mismatchPairs( // returns x-xcenter, y, strength (sign same as disparity)
if (tile_lma_debug_level > -1) {
System.out.println("debug12348973590");
}
double [] mismatch_result;
if (imgdtt_params.ly_poly) {
mismatch_result = corr2d.mismatchPairs( // returns x-xcenter, y, strength (sign same as disparity)
imgdtt_params, // ImageDttParameters imgdtt_params,
corrs, // double [][] corrs,
all_pairs, // int pair_mask, // which pairs to process
-disparity, // double xcenter, // preliminary center x in pixels for largest baseline
imgdtt_params.ortho_vasw_pwr, // double vasw_pwr, // value as weight to this power,
max_corr_radius, // double vasw_pwr, // value as weight to this power,
tile_lma_debug_level,// int debug_level,
tileX, // int tileX, // just for debug output
tileY ); // int tileY
} else {
mismatch_result = corr2d.mismatchPairsCM( // returns x-xcenter, y, strength (sign same as disparity)
imgdtt_params, // ImageDttParameters imgdtt_params,
corrs, // double [][] corrs,
all_pairs, // int pair_mask, // which pairs to process
-disparity, // double xcenter, // preliminary center x in pixels for largest baseline
imgdtt_params.ortho_vasw_pwr, // radius, // positive - within that distance, negative - within 2*(-radius)+1 square
tile_lma_debug_level,// int debug_level,
tileX, // int tileX, // just for debug output
tileY ); // int tileY
}
if (tile_lma_debug_level > 0) {
System.out.println("Lazy eye mismatch:");
for (int np = 0; np < mismatch_result.length/3; np++) {
......@@ -2241,8 +2268,11 @@ public class ImageDtt {
else if (corr_mode == 3) extra_disparity = disparity_map[DISPARITY_INDEX_HOR][tIndex];
else if (corr_mode == 4) extra_disparity = disparity_map[DISPARITY_INDEX_VERT][tIndex];
if (Double.isNaN(extra_disparity)) extra_disparity = 0;
} // if (disparity_map != null){ // not null - calculate correlations
if (Double.isNaN(disparity_map[DISPARITY_STRENGTH_INDEX][tIndex])) {
System.out.println("BUG: 3. disparity_map[DISPARITY_STRENGTH_INDEX][tIndex] should not be NaN");
}
} // if (disparity_map != null){ // not null - calculate correlations
// only debug is left
......@@ -5617,7 +5647,7 @@ public class ImageDtt {
if (i == (transform_size-1)) enh_ortho_scale[i] = 0.0 ; // hardwired 0 in the center
enh_ortho_scale[i] *= Math.sin(Math.PI*(i+1.0)/(2*transform_size));
}
if (globalDebugLevel > 0){
if (globalDebugLevel > 1){
System.out.println("enhortho_width="+ imgdtt_params.enhortho_width+" enhortho_scale="+ imgdtt_params.enhortho_scale);
for (int i = 0; i < corr_size; i++){
System.out.println(" enh_ortho_scale["+i+"]="+ enh_ortho_scale[i]);
......@@ -5797,7 +5827,7 @@ public class ImageDtt {
corr2d.createOrtoNotch(
imgdtt_params.enhortho_width, // double enhortho_width,
imgdtt_params.enhortho_scale, //double enhortho_scale,
true); // boolean debug);
false); // true); // boolean debug);
// public int enhortho_width = 2; // reduce weight of center correlation pixels from center (0 - none, 1 - center, 2 +/-1 from center)
// public double enhortho_scale = 0.0; // 0.2; // multiply center correlation pixels (inside enhortho_width)
......
src/main/java/ImageDttParameters.java
View file @ b6994b6e
......@@ -44,6 +44,7 @@ public class ImageDttParameters {
public int enhortho_width = 2; // reduce weight of center correlation pixels from center (0 - none, 1 - center, 2 +/-1 from center)
public double enhortho_scale = 0.0; // 0.2; // multiply center correlation pixels (inside enhortho_width)
public boolean ly_poly = false; // Use polynomial when measuring mismatch (false - use center of mass)
public boolean fo_correct = true; // correct far objects by comparing orthogonal correlations
public boolean fo_far = false;// Use smallest disparity (false - largest)
......@@ -122,7 +123,9 @@ public class ImageDttParameters {
"Raise value to this power and apply as weight. Reduce width to 3 samples if false, 5 OK when true");
gd.addNumericField("Reduce weight of center correlation pixels from center (0 - none, 1 - center, 2 +/-1 from center)", this.enhortho_width, 0);
gd.addNumericField("Multiply center correlation pixels (inside enhortho_width) (1.0 - disables enh_orttho)", this.enhortho_scale, 3);
gd.addNumericField("Multiply center correlation pixels (inside enhortho_width) (1.0 - disables enh_ortho)", this.enhortho_scale, 3);
gd.addCheckbox ("Use polynomial when measuring mismatch (false - use center of mass)", this.ly_poly);
gd.addMessage("Far objects correction");
gd.addCheckbox ("Try to correct far objects (make them closer) by hor/vert comparison", this.fo_correct);
......@@ -247,6 +250,8 @@ public class ImageDttParameters {
this.enhortho_width= (int) gd.getNextNumber();
this.enhortho_scale= gd.getNextNumber();
this.ly_poly = gd.getNextBoolean();
this.fo_correct = gd.getNextBoolean();
this.fo_far = gd.getNextBoolean();
this.fo_min_strength = gd.getNextNumber();
......@@ -330,6 +335,8 @@ public class ImageDttParameters {
properties.setProperty(prefix+"twice_diagonal", this.twice_diagonal +"");
properties.setProperty(prefix+"min_corr", this.min_corr +"");
properties.setProperty(prefix+"ly_poly", this.ly_poly +"");
properties.setProperty(prefix+"fo_correct", this.fo_correct +"");
properties.setProperty(prefix+"fo_far", this.fo_far +"");
properties.setProperty(prefix+"fo_min_strength", this.fo_min_strength +"");
......@@ -402,6 +409,8 @@ public class ImageDttParameters {
if (properties.getProperty(prefix+"enhortho_scale")!=null) this.enhortho_scale=Double.parseDouble(properties.getProperty(prefix+"enhortho_scale"));
if (properties.getProperty(prefix+"fo_correct")!=null) this.fo_correct=Boolean.parseBoolean(properties.getProperty(prefix+"fo_correct"));
if (properties.getProperty(prefix+"ly_poly")!=null) this.ly_poly=Boolean.parseBoolean(properties.getProperty(prefix+"ly_poly"));
if (properties.getProperty(prefix+"fo_far")!=null) this.fo_far=Boolean.parseBoolean(properties.getProperty(prefix+"fo_far"));
if (properties.getProperty(prefix+"fo_min_strength")!=null) this.fo_min_strength=Double.parseDouble(properties.getProperty(prefix+"fo_min_strength"));
if (properties.getProperty(prefix+"fo_min_eff")!=null) this.fo_min_eff=Double.parseDouble(properties.getProperty(prefix+"fo_min_eff"));
......@@ -481,6 +490,8 @@ public class ImageDttParameters {
idp.enhortho_width = this.enhortho_width;
idp.enhortho_scale = this.enhortho_scale;
idp.ly_poly = this.ly_poly;
idp.fo_correct = this.fo_correct;
idp.fo_far = this.fo_far;
idp.fo_min_strength = this.fo_min_strength;
......
src/main/java/QuadCLT.java
View file @ b6994b6e
......@@ -3807,7 +3807,8 @@ public class QuadCLT {
// visualize correlation results
if (clt_corr_combo!=null){
if (disparity_map != null){
if (!batch_mode && !infinity_corr && clt_parameters.show_map && (debugLevel > -1)){
// if (!batch_mode && !infinity_corr && clt_parameters.show_map && (debugLevel > -1)){
if (!batch_mode && clt_parameters.show_map && (debugLevel > -1)){
sdfa_instance.showArrays(
disparity_map,
tilesX,
......@@ -3821,14 +3822,32 @@ public class QuadCLT {
if (infinity_corr && (disparity_map != null)){
System.out.println("=== applying geometry correction coefficients to correct disparity at infinity ===");
System.out.println("=== Set inf_repeat =0 to disable automatic correction and just generate a data image to correct in offline mode ===");
double [][] inf_ds = {
double [] mismatch_strength = disparity_map[ImageDtt.DISPARITY_STRENGTH_INDEX].clone();
int [] strength_indices = {2,5,8,11};
for (int nt = 0; nt < mismatch_strength.length; nt++) {
if (Double.isNaN(mismatch_strength[nt])) {
mismatch_strength[nt] = 0.0;
} else {
if (mismatch_strength[nt] > 0) {
for (int i = 0; i < strength_indices.length; i++) {
if (!(mismatch_strength[nt] < clt_mismatch[strength_indices[i]][nt])) {
mismatch_strength[nt] = clt_mismatch[strength_indices[i]][nt];
}
// if (nt == 37005) {
// System.out.println("breakpoint in processCLTQuadCorr()");
// }
}
}
}
}
double [][] inf_ds = { // when using with disparity, programmed disparity should be 0
disparity_map[ImageDtt.DISPARITY_INDEX_CM],
disparity_map[ ImageDtt.DISPARITY_STRENGTH_INDEX]};
mismatch_strength}; // disparity_map[ ImageDtt.DISPARITY_STRENGTH_INDEX]};
String [] titles = {"disp_cm", "strength"};
if (!batch_mode && (clt_mismatch != null)){
double [][] inf_ds1 = {
disparity_map[ImageDtt.DISPARITY_INDEX_CM],
disparity_map[ ImageDtt.DISPARITY_STRENGTH_INDEX],
mismatch_strength, //disparity_map[ ImageDtt.DISPARITY_STRENGTH_INDEX],
clt_mismatch[0], // dx0
clt_mismatch[1], // dy0 +
clt_mismatch[3], // dx1
......@@ -3895,7 +3914,7 @@ public class QuadCLT {
clt_parameters.corr_magic_scale, // double magic_coeff, // still not understood coefficent that reduces reported disparity value. Seems to be around 8.5
debugLevel + (clt_parameters.fine_dbg ? 1:0)); // int debugLevel)
if (debugLevel > -1){
if ((new_corr != null) && (debugLevel > -1)){
System.out.println("process_infinity_corr(): ready to apply infinity correction");
show_fine_corr(
new_corr, // double [][][] corr,
......@@ -3965,7 +3984,7 @@ public class QuadCLT {
for (int iQuad = 0; iQuad < clt_data.length; iQuad++){
String title=name+"-"+String.format("%02d", iQuad);
// String titleFull=title+"-SPLIT-D"+clt_parameters.disparity;
// String titleFull=title+"-SPLIT-D"+clt_parameters.disparity;
if (clt_parameters.corr_sigma > 0){ // no filter at all
for (int chn = 0; chn < clt_data[iQuad].length; chn++) {
......@@ -3978,8 +3997,6 @@ public class QuadCLT {
}
}
// int tp.tilesY = imp_quad[iQuad].getHeight()/clt_parameters.transform_size;
// int tp.tilesX = imp_quad[iQuad].getWidth()/clt_parameters.transform_size;
if (debugLevel > 0){
System.out.println("--tilesX="+tilesX);
System.out.println("--tilesY="+tilesY);
......@@ -4059,7 +4076,7 @@ public class QuadCLT {
}
//imp_stack.getProcessor().resetMinAndMax();
//imp_stack.show();
// eyesisCorrections.saveAndShow(imp_stack, this.correctionsParameters);
// eyesisCorrections.saveAndShow(imp_stack, this.correctionsParameters);
eyesisCorrections.saveAndShowEnable(
imp_stack, // ImagePlus imp,
this.correctionsParameters, // EyesisCorrectionParameters.CorrectionParameters correctionsParameters,
......@@ -4068,9 +4085,9 @@ public class QuadCLT {
}
if (clt_parameters.gen_4_img) {
// Save as individual JPEG images in the model directory
String x3d_path= correctionsParameters.selectX3dDirectory(
name, // quad timestamp. Will be ignored if correctionsParameters.use_x3d_subdirs is false
correctionsParameters.x3dModelVersion,
String x3d_path= correctionsParameters.selectX3dDirectory(
name, // quad timestamp. Will be ignored if correctionsParameters.use_x3d_subdirs is false
correctionsParameters.x3dModelVersion,
true, // smart,
true); //newAllowed, // save
for (int sub_img = 0; sub_img < 4; sub_img++){
......@@ -4855,6 +4872,7 @@ public class QuadCLT {
double [][][] new_corr = ac.lazyEyeCorrection(
clt_parameters.ly_poly, // final boolean use_poly,
true, // final boolean restore_disp_inf, // Restore subtracted disparity for scan #0 (infinity)
clt_parameters.fcorr_radius, // final double fcorr_radius,
clt_parameters.fcorr_inf_strength, // final double min_strenth,
clt_parameters.fcorr_inf_diff, // final double max_diff,
......@@ -4960,7 +4978,7 @@ public class QuadCLT {
return results;
}
public void process_infinity_corr(
public void process_infinity_corr( //from existing image
EyesisCorrectionParameters.CLTParameters clt_parameters,
int debugLevel
) {
......@@ -5059,6 +5077,7 @@ public class QuadCLT {
double [][][] new_corr = ac.lazyEyeCorrection(
clt_parameters.ly_poly, // final boolean use_poly,
true, // final boolean restore_disp_inf, // Restore subtracted disparity for scan #0 (infinity)
clt_parameters.fcorr_radius, // final double fcorr_radius,
clt_parameters.fcorr_inf_strength, // final double min_strenth,
clt_parameters.fcorr_inf_diff, // final double max_diff,
......@@ -6037,6 +6056,7 @@ public class QuadCLT {
int bg_scan = 0;
int combo_scan= tp.clt_3d_passes.size()-1;
if (!batch_mode && clt_parameters.show_extrinsic && (debugLevel >-1)) {
// if (!batch_mode && (debugLevel >-1)) {
tp.showScan(
tp.clt_3d_passes.get(bg_scan), // CLTPass3d scan,
"bg_scan"); //String title)
......@@ -6179,7 +6199,7 @@ public class QuadCLT {
double [] dbg_bg_use = new double [bg_sel.length];
double [] dbg_combo_use = new double [bg_sel.length];
for (int i= 0; i < bg_sel.length; i++) {
dbg_bg_sel[i] = bg_sel[i]? 1.0:0.0;
dbg_bg_sel[i] = bg_sel[i]? 1.0:0.0; //only sky, no far mountains (too high disparity!)
dbg_bg_use[i] = bg_use[i]? 1.0:0.0;
dbg_combo_use[i] = combo_use[i]? 1.0:0.0;
}
......@@ -6189,7 +6209,7 @@ public class QuadCLT {
filtered_combo_scand_isp_strength[0],
filtered_combo_scand_isp_strength[1],
dbg_bg_sel,
dbg_bg_use,
dbg_bg_use, // too few
dbg_combo_use};
(new showDoubleFloatArrays()).showArrays(dbg_img, tp.getTilesX(), tp.getTilesY(), true, "extrinsics_bgnd_combo",titles);
}
......@@ -6233,7 +6253,9 @@ public class QuadCLT {
scans14[14 * 0 + 2 + i] = bg_mismatch[i];
scans14[14 * 1 + 2 + i] = combo_mismatch[i];
}
// (new showDoubleFloatArrays()).showArrays(scans14, tp.getTilesX(), tp.getTilesY(), true, "scans_14"); // , titles);
if (debugLevelInner > 0) {
(new showDoubleFloatArrays()).showArrays(scans14, tp.getTilesX(), tp.getTilesY(), true, "scans_14"); // , titles);
}
if (!batch_mode && clt_parameters.show_extrinsic && (debugLevel > 1)) {
tp.showScan(
......@@ -6247,6 +6269,7 @@ public class QuadCLT {
double [][] target_disparity = {tp.clt_3d_passes.get(bg_scan).getDisparity(0), tp.clt_3d_passes.get(combo_scan).getDisparity(0)};
double [][][] new_corr = ac.lazyEyeCorrection(
adjust_poly, // final boolean use_poly,
true, // final boolean restore_disp_inf, // Restore subtracted disparity for scan #0 (infinity)
clt_parameters.fcorr_radius, // final double fcorr_radius,
clt_parameters.fcorr_inf_strength, // final double min_strenth,
clt_parameters.fcorr_inf_diff, // final double max_diff,
......
src/main/java/TileProcessor.java
View file @ b6994b6e
......@@ -656,7 +656,7 @@ public class TileProcessor {
final boolean usePoly, // use polynomial method to find max), valid if useCombo == false
final int debugLevel)
{
final int dbg_tile = (debugLevel > 0)? 839: -1; // x = 122, y= 108; -1; // 27669;
final int dbg_tile = (debugLevel > 0)? -839: -1; // x = 122, y= 108; -1; // 27669;
CLTPass3d combo_pass =new CLTPass3d(this);
final int disparity_index = usePoly ? ImageDtt.DISPARITY_INDEX_POLY : ImageDtt.DISPARITY_INDEX_CM;
combo_pass.tile_op = new int [tilesY][tilesX];
......@@ -3514,12 +3514,15 @@ public class TileProcessor {
{
// final double super_trust = 1.6; // If strength exceeds ex_strength * super_trust, do not apply ex_nstrength
final int dbg_tile = 41030; // x = 206, y = 126 (324*126+206)
final int dbg_tile = -41030; // x = 206, y = 126 (324*126+206)
final boolean show_scan = show_filter_scan || (debugLevel > 1);
showDoubleFloatArrays sdfa_instance = null;
if ((debugLevel > -2) && ((debugLevel > -1) || show_scan)) sdfa_instance = new showDoubleFloatArrays(); // just for debugging?
if (debugLevel > -2){
System.out.println("FilterScan(,,"+disparity_far+", " +disparity_near+", "+ sure_smth);
if (debugLevel > -2){ //-2
if (debugLevel > -1){ //-2
System.out.print("FilterScan(,,"+disparity_far+", " +disparity_near+", "+ sure_smth+"... ");
}
System.out.print(".");
}
final int tlen = tilesY * tilesX;
double [] this_disparity = scan.getDisparity(); // currently calculated, including ortho
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment