Skip to content

I
imagej-elphel
Commit 651267f2 authored by Andrey Filippov's avatar Andrey Filippov
Browse files
Options

Added variable-size 2D correlation output

parent 84eeaf35
Show whitespace changes
Inline Side-by-side
Showing with 109 additions and 54 deletions
src/main/java/com/elphel/imagej/cameras/CLTParameters.java
View file @ 651267f2
......@@ -772,6 +772,7 @@ public class CLTParameters {
public double gpu_sigma_b = 1.1;
public double gpu_sigma_g = 0.7;
public double gpu_sigma_m = 0.7;
public double gpu_sigma_rb_corr = 0.5; // apply LPF after accumulating R and B correlation before G,
public double gpu_sigma_corr = 0.9;
public double gpu_sigma_corr_m = 0.15;
public double gpu_fatz = 30.0;
......@@ -837,6 +838,9 @@ public class CLTParameters {
return monochrome ? gpu_sigma_corr_m : gpu_sigma_corr;
}
public double getGpuCorrRBSigma(boolean monochrome) {
return monochrome ? 1.0 : gpu_sigma_rb_corr;
}
public double getScaleStrength(boolean aux) {
return aux ? scale_strength_aux : scale_strength_main;
......@@ -1540,6 +1544,7 @@ public class CLTParameters {
properties.setProperty(prefix+"gpu_sigma_b", this.gpu_sigma_b +"");
properties.setProperty(prefix+"gpu_sigma_g", this.gpu_sigma_g +"");
properties.setProperty(prefix+"gpu_sigma_m", this.gpu_sigma_m +"");
properties.setProperty(prefix+"gpu_sigma_rb_corr", this.gpu_sigma_rb_corr +"");
properties.setProperty(prefix+"gpu_sigma_corr", this.gpu_sigma_corr +"");
properties.setProperty(prefix+"gpu_sigma_corr_m", this.gpu_sigma_corr_m +"");
properties.setProperty(prefix+"gpu_fatz", this.gpu_fatz +"");
......@@ -2305,6 +2310,7 @@ public class CLTParameters {
if (properties.getProperty(prefix+"gpu_sigma_b")!=null) this.gpu_sigma_b=Double.parseDouble(properties.getProperty(prefix+"gpu_sigma_b"));
if (properties.getProperty(prefix+"gpu_sigma_g")!=null) this.gpu_sigma_g=Double.parseDouble(properties.getProperty(prefix+"gpu_sigma_g"));
if (properties.getProperty(prefix+"gpu_sigma_m")!=null) this.gpu_sigma_m=Double.parseDouble(properties.getProperty(prefix+"gpu_sigma_m"));
if (properties.getProperty(prefix+"gpu_sigma_rb_corr")!=null) this.gpu_sigma_rb_corr=Double.parseDouble(properties.getProperty(prefix+"gpu_sigma_rb_corr"));
if (properties.getProperty(prefix+"gpu_sigma_corr")!=null) this.gpu_sigma_corr=Double.parseDouble(properties.getProperty(prefix+"gpu_sigma_corr"));
if (properties.getProperty(prefix+"gpu_sigma_corr_m")!=null) this.gpu_sigma_corr_m=Double.parseDouble(properties.getProperty(prefix+"gpu_sigma_corr_m"));
if (properties.getProperty(prefix+"gpu_fatz")!=null) this.gpu_fatz=Double.parseDouble(properties.getProperty(prefix+"gpu_fatz"));
......@@ -3211,6 +3217,7 @@ public class CLTParameters {
"Weight of R for composite 2D correlation (green weight is 1.0 -gpu_weight_r - gpu_weight_b");
gd.addNumericField("Correlation weight B", this.gpu_weight_b, 4, 6,"",
"Weight of R for composite 2D correlation (green weight is 1.0 -gpu_weight_r - gpu_weight_b");
gd.addMessage ("--- LPF sigmas used for images/textures ---");
gd.addNumericField("Color LPF sigma R", this.gpu_sigma_r, 4, 6,"pix",
"LPF sigma to process color components during aberration correction");
gd.addNumericField("Color LPF sigma B", this.gpu_sigma_b, 4, 6,"pix",
......@@ -3219,6 +3226,9 @@ public class CLTParameters {
"LPF sigma to process color components during aberration correction");
gd.addNumericField("Monochrome LPF sigma", this.gpu_sigma_m, 4, 6,"pix",
"LPF sigma to process monochrome (e.g.LWIR) during aberration correction");
gd.addMessage ("--- LPF sigmas used for 2D phase correlation ---");
gd.addNumericField("LPF sigma Red/Blue (extra)", this.gpu_sigma_rb_corr, 4, 6,"pix",
"LPF sigma to apply to the composite R+B 2D correlation before adding G (next LPF will be applied to R+B+G correlation");
gd.addNumericField("LPF sigma for correlation, color", this.gpu_sigma_corr, 4, 6,"pix",
"LPF sigma to apply to the composite 2D correlation for RGB images");
gd.addNumericField("LPF sigma for correlation, mono", this.gpu_sigma_corr_m, 4, 6,"pix",
......@@ -3964,6 +3974,7 @@ public class CLTParameters {
this.gpu_sigma_b = gd.getNextNumber();
this.gpu_sigma_g = gd.getNextNumber();
this.gpu_sigma_m = gd.getNextNumber();
this.gpu_sigma_rb_corr = gd.getNextNumber();
this.gpu_sigma_corr = gd.getNextNumber();
this.gpu_sigma_corr_m = gd.getNextNumber();
this.gpu_fatz = gd.getNextNumber();
......
src/main/java/com/elphel/imagej/gpu/GPUTileProcessor.java
View file @ 651267f2
......@@ -765,7 +765,8 @@ public class GPUTileProcessor {
public void execCorr2D(
double [] scales,
double fat_zero) {
double fat_zero,
int corr_radius) {
if (GPU_CORRELATE2D_kernel == null)
{
IJ.showMessage("Error", "No GPU kernel: GPU_CORRELATE2D_kernel");
......@@ -788,6 +789,7 @@ public class GPUTileProcessor {
Pointer.to(new int[] { num_corr_tiles }), // lpf_mask
Pointer.to(gpu_corr_indices),
Pointer.to(new int[] { corr_stride }),
Pointer.to(new int[] { corr_radius }),
Pointer.to(gpu_corrs) // lpf_mask
);
cuCtxSynchronize();
......@@ -800,9 +802,9 @@ public class GPUTileProcessor {
cuCtxSynchronize();
}
public float [][] getCorr2D(){
float [] cpu_corrs = new float [ num_corr_tiles * CORR_SIZE];
public float [][] getCorr2D(int corr_rad){
int corr_size = (2 * corr_rad + 1) * (2 * corr_rad + 1);
float [] cpu_corrs = new float [ num_corr_tiles * corr_size];
CUDA_MEMCPY2D copyD2H = new CUDA_MEMCPY2D();
copyD2H.srcMemoryType = CUmemorytype.CU_MEMORYTYPE_DEVICE;
copyD2H.srcDevice = gpu_corrs;
......@@ -810,16 +812,16 @@ public class GPUTileProcessor {
copyD2H.dstMemoryType = CUmemorytype.CU_MEMORYTYPE_HOST;
copyD2H.dstHost = Pointer.to(cpu_corrs);
copyD2H.dstPitch = CORR_SIZE * Sizeof.FLOAT;
copyD2H.dstPitch = corr_size * Sizeof.FLOAT;
copyD2H.WidthInBytes = CORR_SIZE * Sizeof.FLOAT;
copyD2H.WidthInBytes = corr_size * Sizeof.FLOAT;
copyD2H.Height = num_corr_tiles;
cuMemcpy2D(copyD2H); // run copy
float [][] corrs = new float [num_corr_tiles][ CORR_SIZE];
float [][] corrs = new float [num_corr_tiles][ corr_size];
for (int ncorr = 0; ncorr < num_corr_tiles; ncorr++) {
System.arraycopy(cpu_corrs, ncorr*CORR_SIZE, corrs[ncorr], 0, CORR_SIZE);
System.arraycopy(cpu_corrs, ncorr*corr_size, corrs[ncorr], 0, corr_size);
}
return corrs;
}
......
src/main/java/com/elphel/imagej/tileprocessor/ImageDtt.java
View file @ 651267f2
......@@ -9113,11 +9113,15 @@ public class ImageDtt {
final int transform_len = transform_size*transform_size;
final double afat_zero2 = clt_parameters.getGpuFatZero(is_mono); // ? clt_parameters.gpu_fatz_m : clt_parameters.gpu_fatz;
final double sigma_corr = clt_parameters.getGpuCorrSigma(is_mono); // ? clt_parameters.gpu_sigma_corr_m : clt_parameters.gpu_sigma_corr;
final double sigma_rb_corr = clt_parameters.getGpuCorrRBSigma(is_mono);
final int corr_radius = clt_parameters.gpu_corr_rad;
// final double[] lpf = getLpf(sigma_corr);
final double[] lpf_fd = doubleGetCltLpfFd(
sigma_corr);
final double[] lpf_rb_fd = doubleGetCltLpfFd(
sigma_rb_corr);
for (int ithread = 0; ithread < threads.length; ithread++) {
threads[ithread] = new Thread() {
@Override
......@@ -9199,6 +9203,35 @@ public class ImageDtt {
}
}
}
//c
if (c == 1) { // after red and blue, color only
if (lpf_rb_fd != null) {
for (int n = 0; n<4; n++) {
for (int i = 0; i < transform_len; i++) {
tcorr[n][i] *= lpf_rb_fd[i];
}
}
if (debug_gpu && debug_tile) {
System.out.println("=== LPF-RB for CORRELATION ===");
for (int i = 0; i < transform_size; i++) {
for (int j = 0; j < transform_size; j++) {
System.out.print(String.format("%10.5f ", lpf_rb_fd[transform_size * i + j]));
}
System.out.println();
}
System.out.println("=== R+B LPF-ed CORRELATION ===");
for (int dct_mode = 0; dct_mode < 4; dct_mode++) {
System.out.println("------dct_mode="+dct_mode);
for (int i = 0; i < transform_size; i++) {
for (int j = 0; j < transform_size; j++) {
System.out.print(String.format("%10.3f ", tcorr[dct_mode][transform_size * i + j]));
}
System.out.println();
}
}
}
}
}
}
}
corrs2d[n_corr] = correlation2d.normalizeConvertCorr(
......
src/main/java/com/elphel/imagej/tileprocessor/TwoQuadCLT.java
View file @ 651267f2
......@@ -1994,7 +1994,8 @@ public class TwoQuadCLT {
for (int i = 0; i < NREPEAT; i++ ) gPUTileProcessor.execCorr2D(
scales,// double [] scales,
fat_zero); // double fat_zero);
fat_zero, // double fat_zero);
clt_parameters.gpu_corr_rad); // int corr_radius
long endCorr2d = System.nanoTime();
......@@ -2037,12 +2038,13 @@ public class TwoQuadCLT {
debugLevel );
}
float [][] corr2D = gPUTileProcessor.getCorr2D();
float [][] corr2D = gPUTileProcessor.getCorr2D(
clt_parameters.gpu_corr_rad); // int corr_rad);
// convert to 6-layer image using tasks
int tilesX = GPUTileProcessor.IMG_WIDTH / GPUTileProcessor.DTT_SIZE;
int tilesY = GPUTileProcessor.IMG_HEIGHT / GPUTileProcessor.DTT_SIZE;
int [] wh = new int[2];
double [][] dbg_corr = gPUTileProcessor.getCorr2DView(
double [][] dbg_corr = GPUTileProcessor.getCorr2DView(
tilesX,
tilesY,
corr_indices,
......@@ -2054,7 +2056,7 @@ public class TwoQuadCLT {
wh[1],
true,
"CORR2D",
gPUTileProcessor.getCorrTitles());
GPUTileProcessor.getCorrTitles());
if (clt_parameters.gen_chn_img) {
// combine to a sliced color image
......
src/main/resources/TileProcessor.cuh
View file @ 651267f2
......@@ -57,7 +57,8 @@
#define IMCLT_TILES_PER_BLOCK 4
#define CORR_PAIR_SHIFT 8 // 8 lower bits - number of a pair, other bits tile number
#define TASK_CORR_BITS 4
#define CORR_OUT_RAD 7
#define CORR_OUT_RAD 4
//7
#endif
......@@ -376,6 +377,7 @@ __device__ void debug_print_mclt(
float * mclt_tile, // [4][DTT_SIZE][DTT_SIZE1], // +1 to alternate column ports)
const int color);
__device__ void debug_print_corr_15x15(
int corr_radius,
float * mclt_tile, //DTT_SIZE2M1 x DTT_SIZE2M1
const int color);
// Fractional pixel shift (phase rotation), horizontal. In-place.
......@@ -400,6 +402,7 @@ __device__ void normalizeTileAmplitude(
float * clt_tile, // [4][DTT_SIZE][DTT_SIZE1], // +1 to alternate column ports
float fat_zero); // fat zero is absolute, scale it outside
__device__ void corrUnfoldTile(
int corr_radius,
float* qdata0, // [4][DTT_SIZE][DTT_SIZE1], // 4 quadrants of the clt data, rows extended to optimize shared ports
float* rslt); // [DTT_SIZE2M1][DTT_SIZE2M1]) // 15x15
__device__ void imclt( // implemented, used // why is it twice?
......@@ -426,6 +429,7 @@ __global__ void correlate2D(
size_t num_corr_tiles, // number of correlation tiles to process
int * gpu_corr_indices, // packed tile+pair
const size_t corr_stride, // in floats
int corr_radius, // radius of the output correlation (7 for 15x15)
float * gpu_corrs) // correlation output data
{
/// int thr3 = threadIdx.x >> 3; // now zero?
......@@ -559,27 +563,17 @@ __global__ void correlate2D(
#endif
#endif
// now new part - need to transform with DCT-II and make 15x15
/*
// quadrant 0 dct_ii hor, dct_ii vert,
// quadrant 1 dct_ii hor, dst_ii vert,
// quadrant 2 dst_ii hor, dct_ii vert,
// quadrant 3 dst_ii hor, dst_ii vert,
Java code:
for (int quadrant = 0; quadrant < 4; quadrant++){
int mode = ((quadrant << 1) & 2) | ((quadrant >> 1) & 1); // transpose
tcorr[first_col][quadrant] = dtt.dttt_iie(tcorr[first_col][quadrant], mode, transform_size);
}
*/
// change to 16-32 threads?? in next iteration
// hor pass
// vert pass (hor pass in Java, before transpose. Here transposed, no transform needed)
for (int q = 0; q < 4; q++){
int is_sin = (q >> 1) & 1;
// int is_sin = q & 1;
// dttii_shared_mem(clt_corr + (q * DTT_SIZE + threadIdx.x) * DTT_SIZE1 , 1, is_sin); // horizontal pass, tread is row
// dttii_shared_mem(clt_corr + q * (DTT_SIZE1 * DTT_SIZE) + threadIdx.x , DTT_SIZE1, is_sin); // vertical pass, thread is column
dttii_shared_mem_nonortho(clt_corr + q * (DTT_SIZE1 * DTT_SIZE) + threadIdx.x , DTT_SIZE1, is_sin); // vertical pass, thread is column
}
__syncthreads();
......@@ -593,26 +587,25 @@ Java code:
#endif
#endif
// vert pass
// hor pass, corresponding to vert pass in Java
for (int q = 0; q < 4; q++){
int is_sin = q & 1;
// int is_sin = (q >> 1) & 1;
// dttii_shared_mem(clt_corr + q * (DTT_SIZE1 * DTT_SIZE) + threadIdx.x , DTT_SIZE1, is_sin); // vertical pass, thread is column
// dttii_shared_mem(clt_corr + (q * DTT_SIZE + threadIdx.x) * DTT_SIZE1 , 1, is_sin); // horizontal pass, tread is row
dttii_shared_mem_nonortho(clt_corr + (q * DTT_SIZE + threadIdx.x) * DTT_SIZE1 , 1, is_sin); // horizontal pass, tread is row
}
__syncthreads();
#ifdef DBG_TILE
#ifdef DEBUG6
if ((tile_num == DBG_TILE) && (corr_pair == 0) && (threadIdx.x == 0)){
printf("\ncorrelate2D AFTER HOSIZONTAL (VERTICAL) PASS\n");
if ((tile_num == DBG_TILE) && (corr_pair == 0) && (threadIdx.x == 4)){
printf("\ncorrelate2D AFTER HOSIZONTAL (VERTICAL) PASS, corr_radius=%d\n",corr_radius);
debug_print_clt1(clt_corr, -1, 0xf);
}
__syncthreads();// __syncwarp();
#endif
#endif
corrUnfoldTile(
corr_radius, // int corr_radius,
(float *) clt_corr, // float* qdata0, // [4][DTT_SIZE][DTT_SIZE1], // 4 quadrants of the clt data, rows extended to optimize shared ports
(float *) mclt_corr); // float* rslt) // [DTT_SIZE2M1][DTT_SIZE2M1]) // 15x15
......@@ -621,23 +614,28 @@ Java code:
#ifdef DBG_TILE
#ifdef DEBUG6
if ((tile_num == DBG_TILE) && (corr_pair == 0) && (threadIdx.x == 0)){
printf("\ncorrelate2D after UNFOLD\n");
debug_print_corr_15x15(mclt_corr, -1);
printf("\ncorrelate2D after UNFOL, corr_radius=%d\n",corr_radius);
debug_print_corr_15x15(
corr_radius, // int corr_radius,
mclt_corr,
-1);
}
__syncthreads();// __syncwarp();
#endif
#endif
// copy 15x15 tile to main memory
// searching for bug. Uncomment later
// copy 15x15 tile to main memory (2 * corr_radius +1) x (2 * corr_radius +1)
int size2r1 = 2 * corr_radius + 1;
int len2r1x2r1 = size2r1 * size2r1;
int corr_tile_offset = + corr_stride * corr_num;
float *mem_corr = gpu_corrs + corr_tile_offset;
//CORR_THREADS_PER_TILE
// int offs = threadIdx.x;
#pragma unroll
for (int offs = threadIdx.x; offs < DTT_SIZE2M1*DTT_SIZE2M1; offs+=CORR_THREADS_PER_TILE){ // variable number of cycles per thread
// for (int offs = threadIdx.x; offs < DTT_SIZE2M1*DTT_SIZE2M1; offs+=CORR_THREADS_PER_TILE){ // variable number of cycles per thread
for (int offs = threadIdx.x; offs < len2r1x2r1; offs+=CORR_THREADS_PER_TILE){ // variable number of cycles per thread
mem_corr[offs] = mclt_corr[offs];
}
__syncthreads();
#ifdef DBG_TILE
#ifdef DEBUG6
......@@ -969,27 +967,36 @@ Converted from DttRad2.java:443
)
*/
__device__ void corrUnfoldTile(
int corr_radius,
float* qdata0, // [4][DTT_SIZE][DTT_SIZE1], // 4 quadrants of the clt data, rows extended to optimize shared ports
float* rslt) // [DTT_SIZE2M1][DTT_SIZE2M1]) // 15x15
{
const int rslt_base_index = DTT_SIZE2M1 * (DTT_SIZE) - DTT_SIZE;
int size2r1 = 2 * corr_radius + 1; // 15
int crp1 = corr_radius + 1; //8
/// const int rslt_base_index = DTT_SIZE2M1 * (DTT_SIZE) - DTT_SIZE; // offset of the center
int rslt_base_index = size2r1 * crp1 - crp1; // offset of the center
float * qdata1 = qdata0 + (DTT_SIZE * DTT_SIZE1);
float * qdata2 = qdata1 + (DTT_SIZE * DTT_SIZE1);
float * qdata3 = qdata2 + (DTT_SIZE * DTT_SIZE1);
int i = threadIdx.x;
if (i > corr_radius) {
return; // not needed, only use inner
}
// printf("\corrUnfoldTile() corr_radius=%d, i=%d\n",corr_radius,i);
float corr_pixscale = 0.25f;
int i_transform_size = i * DTT_SIZE1; // used to address source rows which are 9 long
int im1_transform_size = i_transform_size - DTT_SIZE1; // negative for i = 0, use only after divergence
int rslt_row_offs = i * DTT_SIZE2M1;
/// int rslt_row_offs = i * DTT_SIZE2M1;
int rslt_row_offs = i * size2r1;
int rslt_base_index_p = rslt_base_index + rslt_row_offs; // i * DTT_SIZE2M1;
int rslt_base_index_m = rslt_base_index - rslt_row_offs; // i * DTT_SIZE2M1;
rslt[rslt_base_index_p] = corr_pixscale * qdata0[i_transform_size]; // incomplete, will only be used for thread i=0
rslt[rslt_base_index_m] = rslt[rslt_base_index_p]; // nop for i=0 incomplete, will only be used for thread i=0
for (int j = 1; j < DTT_SIZE; j++) {
/// for (int j = 1; j < DTT_SIZE; j++) {
for (int j = 1; j <= corr_radius; j++) {
int rslt_base_index_pp = rslt_base_index_p + j;
int rslt_base_index_pm = rslt_base_index_p - j;
/// int rslt_base_index_mp = rslt_base_index_m + j;
/// int rslt_base_index_mm = rslt_base_index_m - j;
rslt[rslt_base_index_pp] = corr_pixscale * (
qdata0[i_transform_size + j] +
qdata1[i_transform_size + j -1]); // incomplete, will only be used for thread i=0
......@@ -1000,12 +1007,11 @@ __device__ void corrUnfoldTile(
if (i == 0) {
return;
}
/// int im1 = i-1;
im1_transform_size = i_transform_size - DTT_SIZE1;
/// im1_transform_size = i_transform_size - DTT_SIZE1; // already is calculated
float d = corr_pixscale * qdata2[im1_transform_size];
rslt[rslt_base_index_p] += d;
rslt[rslt_base_index_m] -= d;
for (int j = 1; j < DTT_SIZE; j++) {
for (int j = 1; j <= corr_radius; j++) {
int rslt_base_index_pp = rslt_base_index_p + j;
int rslt_base_index_pm = rslt_base_index_p - j;
int rslt_base_index_mp = rslt_base_index_m + j;
......@@ -1068,14 +1074,15 @@ __device__ void debug_print_mclt(
}
__device__ void debug_print_corr_15x15(
int corr_radius,
float * mclt_tile, //DTT_SIZE2M1 x DTT_SIZE2M1
const int color)
{
int size2r1 = 2 * corr_radius + 1;
if (color >= 0) printf("----------- Color = %d -----------\n",color);
for (int dbg_row = 0; dbg_row < DTT_SIZE2M1; dbg_row++){
for (int dbg_col = 0; dbg_col < DTT_SIZE2M1; dbg_col++){
printf ("%10.5f ", mclt_tile[dbg_row * DTT_SIZE2M1 + dbg_col]);
for (int dbg_row = 0; dbg_row < size2r1; dbg_row++){
for (int dbg_col = 0; dbg_col < size2r1; dbg_col++){
printf ("%10.5f ", mclt_tile[dbg_row * size2r1 + dbg_col]);
}
printf("\n");
}
......
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