Loading src/main/java/CLTPass3d.java +2 −2 Original line number Diff line number Diff line Loading @@ -36,7 +36,7 @@ public class CLTPass3d{ double [] calc_disparity_hor = null; // composite disparity, calculated from "disparity", and "disparity_map" fields double [] calc_disparity_vert = null; // composite disparity, calculated from "disparity", and "disparity_map" fields double [] calc_disparity_combo = null; // composite disparity, calculated from "disparity", and "disparity_map" fields double [] strength = null; // composite strength, initially uses a copy of raw 4-sensor correleation strength double [] strength = null; // composite strength, initially uses a copy of raw 4-sensor correlation strength double [] strength_hor = null; // updated hor strength, initially uses a copy of raw measured double [] strength_vert = null; // updated hor strength, initially uses a copy of raw measured // Bg disparity & strength is calculated from the supertiles and used instead of the tile disparity if it is too weak. Assuming, that Loading Loading @@ -333,7 +333,7 @@ public class CLTPass3d{ } /** * Get one of the line-scan per-tile correlation data. * @param mode 0 - final data (initially copy FPGA generated 4-pair correation) * @param mode 0 - final data (initially copy FPGA generated 4-pair correlation) * 1 - original FPGA generated 4-sensor correlation * 2 - 2 - horizontal pairs correlation, detecting vertical features * 3 - 2 - vertical pairs correlation, detecting horizontal features Loading src/main/java/EyesisCorrectionParameters.java +10 −1 Original line number Diff line number Diff line Loading @@ -2182,6 +2182,7 @@ public class EyesisCorrectionParameters { public boolean replaceWeakOutlayers = true; // false; public boolean dbg_migrate = true; // other debug images public boolean show_ortho_combine = false; // Show 'ortho_combine' Loading Loading @@ -2452,6 +2453,8 @@ public class EyesisCorrectionParameters { properties.setProperty(prefix+"plSnapDispWeight", this.plSnapDispWeight +""); properties.setProperty(prefix+"plSnapZeroMode", this.plSnapZeroMode+""); properties.setProperty(prefix+"dbg_migrate", this.dbg_migrate+""); properties.setProperty(prefix+"show_ortho_combine", this.show_ortho_combine+""); properties.setProperty(prefix+"show_refine_supertiles", this.show_refine_supertiles+""); properties.setProperty(prefix+"show_bgnd_nonbgnd", this.show_bgnd_nonbgnd+""); Loading Loading @@ -2712,6 +2715,8 @@ public class EyesisCorrectionParameters { if (properties.getProperty(prefix+"plSnapDispWeight")!=null) this.plSnapDispWeight=Double.parseDouble(properties.getProperty(prefix+"plSnapDispWeight")); if (properties.getProperty(prefix+"plSnapZeroMode")!=null) this.plPrecision=Integer.parseInt(properties.getProperty(prefix+"plSnapZeroMode")); if (properties.getProperty(prefix+"dbg_migrate")!=null) this.dbg_migrate=Boolean.parseBoolean(properties.getProperty(prefix+"dbg_migrate")); if (properties.getProperty(prefix+"show_ortho_combine")!=null) this.show_ortho_combine=Boolean.parseBoolean(properties.getProperty(prefix+"show_ortho_combine")); if (properties.getProperty(prefix+"show_refine_supertiles")!=null) this.show_refine_supertiles=Boolean.parseBoolean(properties.getProperty(prefix+"show_refine_supertiles")); if (properties.getProperty(prefix+"show_bgnd_nonbgnd")!=null) this.show_bgnd_nonbgnd=Boolean.parseBoolean(properties.getProperty(prefix+"show_bgnd_nonbgnd")); Loading Loading @@ -2995,6 +3000,8 @@ public class EyesisCorrectionParameters { gd.addNumericField("Maximal disparity diff. by weight product to snap to plane", this.plSnapDispWeight, 6); gd.addNumericField("Zero strength snap mode: 0: no special treatment, 1 - strongest, 2 - farthest",this.plSnapZeroMode, 0); gd.addCheckbox ("Test new mode after migration", this.dbg_migrate); gd.addMessage ("--- Other debug images ---"); gd.addCheckbox ("Show 'ortho_combine'", this.show_ortho_combine); gd.addCheckbox ("Show 'refine_disparity_supertiles'", this.show_refine_supertiles); Loading Loading @@ -3264,6 +3271,8 @@ public class EyesisCorrectionParameters { this.plSnapDispWeight= gd.getNextNumber(); this.plSnapZeroMode= (int) gd.getNextNumber(); this.dbg_migrate= gd.getNextBoolean(); this.show_ortho_combine= gd.getNextBoolean(); this.show_refine_supertiles=gd.getNextBoolean(); this.show_bgnd_nonbgnd= gd.getNextBoolean(); // first on second pass Loading src/main/java/MeasuredLayers.java 0 → 100644 +513 −0 Original line number Diff line number Diff line /** ** MeasuredLayer - per-tile measured disparity/strength pairs, ** multiple layers can be used for 4-disparity and 2 (hor/vert) pairs ** separately ** ** Copyright (C) 2017 Elphel, Inc. ** ** -----------------------------------------------------------------------------** ** ** MeasuredLayer.java is free software: you can redistribute it and/or modify ** it under the terms of the GNU General Public License as published by ** the Free Software Foundation, either version 3 of the License, or ** (at your option) any later version. ** ** This program is distributed in the hope that it will be useful, ** but WITHOUT ANY WARRANTY; without even the implied warranty of ** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ** GNU General Public License for more details. ** ** You should have received a copy of the GNU General Public License ** along with this program. If not, see <http://www.gnu.org/licenses/>. ** -----------------------------------------------------------------------------** ** */ public class MeasuredLayers { private DispStrength [][] layers; private int tilesX; private int tilesY; private int superTileSize; private double [][] lapWeight = null; public class DispStrength { public double disparity = Double.NaN; public double strength = 0.0; public DispStrength ( double disparity, double strength) { this.disparity = disparity; this.strength = strength; if (Double.isNaN(disparity)){ this.strength = 0.0; } } public double [] getDispStrength() { double [] ds = {disparity, strength}; return ds; } public void setDisparity(double disparity) { this.disparity = disparity; if (Double.isNaN(disparity)){ this.strength = 0.0; } } public void setStrength(double strength) { this.strength = strength; } public double getDisparity() { return disparity; } public double getStrength() { return strength; } } /** * Create a set of measured layers. Several layers can be used to represent alternative measurements * (with different preset disparity) or for different modes (4 pair correlation, vertical, horizontal) * @param num_layers number of alternative layers used (3 for quad, hor, vert) * @param tilesX number of tiles horizontally in the image * @param tilesY number of tiles vertically in the image * @param superTileSize size of the supertile square */ public MeasuredLayers( int num_layers, int tilesX, int tilesY, int superTileSize) { layers = new DispStrength [num_layers][]; this.tilesX = tilesX; this.tilesY = tilesY; this.superTileSize = superTileSize; this.lapWeight = getLapWeights(); } /** * Get number of tiles in the image horizontally * @return tilesX */ public int getTilesX() { return tilesX; } /** * Get number of tiles in the image vertically * @return tilesY */ public int getTilesY() { return tilesY; } /** * Get number of tiles in each supertile in eac direction * @return superTileSize */ public int getSuperTileSize() { return superTileSize; } /** * Set measured layer * @param num_layer number of the layer to set * @param disparity array of per-tile disparity values (linescan order), * NaN deletes the tile * @param strength array of per-tile disparity values (linescan order), * 0.0 strength is OK - does not delete the tile * @param selection optional tile selection (or null). If unselected, tile is deleted */ public void setLayer ( int num_layer, double [] disparity, double [] strength, boolean [] selection) // may be null { if (layers[num_layer] == null) { int llen = disparity.length; if (layers[num_layer] == null) { layers[num_layer] = new DispStrength [llen]; } } for (int i = 0; i < disparity.length; i++){ if ((selection != null) && !selection[i]){ layers[num_layer][i] = null; } else { if (Double.isNaN(disparity[i])){ layers[num_layer][i] = null; } else { if (layers[num_layer][i] != null){ layers[num_layer][i].setDisparity(disparity[i]); layers[num_layer][i].setStrength(strength[i]); } else { layers[num_layer][i] = new DispStrength(disparity[i], strength[i]); } } } } } /** * Set disparity values for selected measurement layer * @param num_layer number of the layer to set * @param disparity array of per-tile disparity values (linescan order), * NaN deletes the tile. New tile set strength = 0.0 */ public void setDisparity ( int num_layer, double [] disparity) { if (layers[num_layer] == null) { int llen = disparity.length; if (layers[num_layer] == null) { layers[num_layer] = new DispStrength [llen]; } } for (int i = 0; i < disparity.length; i++){ if (Double.isNaN(disparity[i])){ layers[num_layer][i] = null; } else { if (layers[num_layer][i] != null){ layers[num_layer][i].setDisparity(disparity[i]); } else { layers[num_layer][i] = new DispStrength(disparity[i], 0.0); } } } } /** * Set strength values for selected measurement layer. Does not allocate new tiles * if that element was null * @param num_layer number of the layer to set * @param strength array of per-tile strength values (linescan order) */ public void setStrength ( int num_layer, double [] strength) { if (layers[num_layer] != null) { for (int i = 0; i < strength.length; i++){ if (layers[num_layer][i] != null){ layers[num_layer][i].setStrength(strength[i]); } } } } /** * Change tile selection for the layer. Can only unselect existing tiles, not add new ones * @param num_layer number of the layer to set * @param selection array of per-tile boolean selection values (linescan order), */ public void setSelection ( int num_layer, boolean [] selection) { for (int i = 0; i < selection.length; i++){ if (!selection[i]){ layers[num_layer][i] = null; } } } /** * Get total number of the measurement layers in this object * @return number of layers (some may be uninitialized */ public int getnumLayers() { if (layers == null){ return 0; } else { return layers.length; } } /** * Get array of disparity values for the selected layer * @param num_layer number of the layer to read * @return array of per-tile disparity values (in linescan order), * Double.NaN for missing tiles */ public double [] getDisparity( int num_layer) { double [] disparity = new double [layers[num_layer].length]; for (int i = 0; i < disparity.length; i++){ if (layers[num_layer][i] == null){ disparity[i] = Double.NaN; } else { disparity[i] = layers[num_layer][i].getDisparity(); } } return disparity; } /** * Get array of correlation strength values for the selected layer * @param num_layer number of the layer to read * @return array of per-tile strength values (in linescan order), * 0.0 for missing tiles */ public double [] getStrength( int num_layer) { double [] strength = new double [layers[num_layer].length]; for (int i = 0; i < strength.length; i++){ if (layers[num_layer][i] == null){ strength[i] = 0.0; } else { strength[i] = layers[num_layer][i].getStrength(); } } return strength; } /** * Get array of existing tiles for the selected layer * @param num_layer number of the layer to read * @return boolean (per tile) array of existing tiles */ public boolean [] getSelection( int num_layer) { boolean [] selection = new boolean [layers[num_layer].length]; for (int i = 0; i < selection.length; i++){ selection[i] = layers[num_layer][i] != null; } return selection; } /** * Calculate weights for overlapping supertiles to multiply correlation strengths * @return square 2-d array of weights 0.0 ... 1.0, each dimension twice the * supertile size */ private double [][] getLapWeights(){ final double [][] lapWeight = new double [2 * superTileSize][2 * superTileSize]; final double [] lapWeight1d = new double [superTileSize]; final int superTileSize2 = 2 * superTileSize; for (int i = 0; i < superTileSize; i++){ lapWeight1d[i] = 0.5*(1.0 - Math.cos((i + 0.5)* Math.PI/superTileSize)); } for (int i = 0; i < superTileSize; i++){ for (int j = 0; j < superTileSize; j++){ lapWeight[i] [ j] = lapWeight1d[i]*lapWeight1d[j]; lapWeight[superTileSize2 - 1 - i][ j] = lapWeight[i][j]; lapWeight[i] [superTileSize2 - 1 - j] = lapWeight[i][j]; lapWeight[superTileSize2 - 1 - i][superTileSize2 - 1 - j] = lapWeight[i][j]; } } double s = 0.0; for (int i = 0; i < superTileSize2; i++){ for (int j = 0; j < superTileSize2; j++){ s+=lapWeight[i][j]; } } System.out.println("getLapWeights: sum = "+s); return lapWeight; } /** * Get selection for the specific measurement layer and supertile X,Y coordinates * in the image. Combined with input selection * @param num_layer number of the measurement layer to process * @param stX supertile horizontal position in the image * @param stY supertile vertical position in the image * @param sel_in optional selection for this supertile (linescan, 4 * supetile size) * @param null_if_none return null if there are no selected tiles in teh result selection * @return boolean array [4*superTileSize] of the selected tiles on the specified * measurement layer */ public boolean [] getSupertileSelection( int num_layer, int stX, int stY, boolean [] sel_in, double strength_floor, boolean null_if_none) { if ((layers[num_layer] == null) && null_if_none){ return null; } int st2 = 2 * superTileSize; int st_half = superTileSize/2; boolean [] selection = new boolean [st2 * st2]; int num_selected = 0; if (layers[num_layer] != null) { for (int dy = 0; dy < st2; dy ++){ int y = superTileSize * stY -st_half + dy; if ((y >= 0) && (y < tilesY)) { for (int dx = 0; dx < st2; dx ++){ int x = superTileSize * stX -st_half + dx; if ((x >= 0) && (x < tilesX)) { int indx = y * tilesX + x; int indx_st = dy * st2 + dx; if (((sel_in == null) || sel_in[indx_st]) && (layers[num_layer][indx] != null) && (layers[num_layer][indx].getStrength() >= strength_floor)){ selection[indx_st] = true; num_selected ++; } } } } } } if (null_if_none && (num_selected == 0)) return null; return selection; } /** * Get selection for the specific measurement layer and supertile X,Y coordinates * in the image and combine with disparity far/near limits. Alse combined with * input selection * @param num_layer number of the measurement layer to process * @param stX supertile horizontal position in the image * @param stY supertile vertical position in the image * @param sel_in optional selection for this supertile (linescan, 4 * supetile size) * @param disp_far lowest acceptable disparity value. Double.NaN - do not check. * @param disp_near highest acceptable disparity value. Double.NaN - do not check. * @param null_if_none return null if there are no selected tiles in teh result selection * @return boolean array [4*superTileSize] of the selected tiles on the specified * measurement layer */ public boolean [] getSupertileSelection( int num_layer, int stX, int stY, boolean [] sel_in, double disp_far, double disp_near, double strength_floor, boolean null_if_none) { if ((layers[num_layer] == null) && null_if_none){ return null; } int st2 = 2 * superTileSize; int st_half = superTileSize/2; boolean [] selection = new boolean [st2 * st2]; int num_selected = 0; if (layers[num_layer] != null) { for (int dy = 0; dy < st2; dy ++){ int y = superTileSize * stY -st_half + dy; if ((y >= 0) && (y < tilesY)) { for (int dx = 0; dx < st2; dx ++){ int x = superTileSize * stX -st_half + dx; if ((x >= 0) && (x < tilesX)) { int indx = y * tilesX + x; int indx_st = dy * st2 + dx; if (((sel_in == null) || sel_in[indx_st]) && (layers[num_layer][indx] != null)){ if ( (Double.isNaN(disp_far) || (layers[num_layer][indx].getDisparity() >= disp_far)) && (Double.isNaN(disp_near) || (layers[num_layer][indx].getDisparity() <= disp_near)) && (layers[num_layer][indx].getStrength() >= strength_floor)){ selection[indx_st] = true; num_selected ++; } } } } } } } if (null_if_none && (num_selected == 0)) return null; return selection; } /** * Get number of "true" elements in a boolean array. Null is OK, it results in 0 * @param selected boolean array to count set elements in * @return number of selected elements */ public static int getNumSelected( boolean [] selected) { if (selected == null) return 0; int num_selected = 0; for (int i = 0; i < selected.length; i++){ if (selected[i]) num_selected ++; } return num_selected; } public static double getSumStrength( double [][] disp_strength) { if (disp_strength == null) return 0.0; double sw = 0.0; for (int i = 0; i < disp_strength[1].length; i++){ sw += disp_strength[1][i]; } return sw; } /** * Get disparity and correlation strength for the specific measurement layer and * supertile X,Y coordinates in the image. Combined with input selection * @param num_layer number of the measurement layer to process * @param stX supertile horizontal position in the image * @param stY supertile vertical position in the image * @param sel_in optional selection for this supertile (linescan, 4 * supetile size) * @param null_if_none return null if there are no selected tiles in the result selection * @param strength_pow Non-linear treatment of correlation strength. Raise data to the strength_pow power * @return double [2][4*superTileSize] {disparity[4*superTileSize], strength [4*superTileSize]} */ public double[][] getDisparityStrength ( int num_layer, int stX, int stY, boolean [] sel_in, double strength_floor, double strength_pow, boolean null_if_none) { if ((layers[num_layer] == null) && null_if_none){ return null; } int st2 = 2 * superTileSize; int st_half = superTileSize/2; double [][] ds = new double [2][st2*st2]; int num_selected = 0; if (layers[num_layer] != null) { for (int dy = 0; dy < st2; dy ++){ int y = superTileSize * stY -st_half + dy; if ((y >= 0) && (y < tilesY)) { for (int dx = 0; dx < st2; dx ++){ int x = superTileSize * stX -st_half + dx; if ((x >= 0) && (x < tilesX)) { int indx = y * tilesX + x; int indx_st = dy * st2 + dx; if (((sel_in == null) || sel_in[indx_st]) && (layers[num_layer][indx] != null)){ ds[0][indx_st] = layers[num_layer][indx].getDisparity(); double w = layers[num_layer][indx].getStrength() - strength_floor; if (w > 0) { if (strength_pow != 1.0) w = Math.pow(w, strength_pow); w *= lapWeight[dy][dx]; ds[0][indx_st] = layers[num_layer][indx].getDisparity(); ds[1][indx_st] = w; num_selected ++; } } } } } } } if (null_if_none && (num_selected == 0)) return null; return ds; } } Loading
src/main/java/CLTPass3d.java +2 −2 Original line number Diff line number Diff line Loading @@ -36,7 +36,7 @@ public class CLTPass3d{ double [] calc_disparity_hor = null; // composite disparity, calculated from "disparity", and "disparity_map" fields double [] calc_disparity_vert = null; // composite disparity, calculated from "disparity", and "disparity_map" fields double [] calc_disparity_combo = null; // composite disparity, calculated from "disparity", and "disparity_map" fields double [] strength = null; // composite strength, initially uses a copy of raw 4-sensor correleation strength double [] strength = null; // composite strength, initially uses a copy of raw 4-sensor correlation strength double [] strength_hor = null; // updated hor strength, initially uses a copy of raw measured double [] strength_vert = null; // updated hor strength, initially uses a copy of raw measured // Bg disparity & strength is calculated from the supertiles and used instead of the tile disparity if it is too weak. Assuming, that Loading Loading @@ -333,7 +333,7 @@ public class CLTPass3d{ } /** * Get one of the line-scan per-tile correlation data. * @param mode 0 - final data (initially copy FPGA generated 4-pair correation) * @param mode 0 - final data (initially copy FPGA generated 4-pair correlation) * 1 - original FPGA generated 4-sensor correlation * 2 - 2 - horizontal pairs correlation, detecting vertical features * 3 - 2 - vertical pairs correlation, detecting horizontal features Loading
src/main/java/EyesisCorrectionParameters.java +10 −1 Original line number Diff line number Diff line Loading @@ -2182,6 +2182,7 @@ public class EyesisCorrectionParameters { public boolean replaceWeakOutlayers = true; // false; public boolean dbg_migrate = true; // other debug images public boolean show_ortho_combine = false; // Show 'ortho_combine' Loading Loading @@ -2452,6 +2453,8 @@ public class EyesisCorrectionParameters { properties.setProperty(prefix+"plSnapDispWeight", this.plSnapDispWeight +""); properties.setProperty(prefix+"plSnapZeroMode", this.plSnapZeroMode+""); properties.setProperty(prefix+"dbg_migrate", this.dbg_migrate+""); properties.setProperty(prefix+"show_ortho_combine", this.show_ortho_combine+""); properties.setProperty(prefix+"show_refine_supertiles", this.show_refine_supertiles+""); properties.setProperty(prefix+"show_bgnd_nonbgnd", this.show_bgnd_nonbgnd+""); Loading Loading @@ -2712,6 +2715,8 @@ public class EyesisCorrectionParameters { if (properties.getProperty(prefix+"plSnapDispWeight")!=null) this.plSnapDispWeight=Double.parseDouble(properties.getProperty(prefix+"plSnapDispWeight")); if (properties.getProperty(prefix+"plSnapZeroMode")!=null) this.plPrecision=Integer.parseInt(properties.getProperty(prefix+"plSnapZeroMode")); if (properties.getProperty(prefix+"dbg_migrate")!=null) this.dbg_migrate=Boolean.parseBoolean(properties.getProperty(prefix+"dbg_migrate")); if (properties.getProperty(prefix+"show_ortho_combine")!=null) this.show_ortho_combine=Boolean.parseBoolean(properties.getProperty(prefix+"show_ortho_combine")); if (properties.getProperty(prefix+"show_refine_supertiles")!=null) this.show_refine_supertiles=Boolean.parseBoolean(properties.getProperty(prefix+"show_refine_supertiles")); if (properties.getProperty(prefix+"show_bgnd_nonbgnd")!=null) this.show_bgnd_nonbgnd=Boolean.parseBoolean(properties.getProperty(prefix+"show_bgnd_nonbgnd")); Loading Loading @@ -2995,6 +3000,8 @@ public class EyesisCorrectionParameters { gd.addNumericField("Maximal disparity diff. by weight product to snap to plane", this.plSnapDispWeight, 6); gd.addNumericField("Zero strength snap mode: 0: no special treatment, 1 - strongest, 2 - farthest",this.plSnapZeroMode, 0); gd.addCheckbox ("Test new mode after migration", this.dbg_migrate); gd.addMessage ("--- Other debug images ---"); gd.addCheckbox ("Show 'ortho_combine'", this.show_ortho_combine); gd.addCheckbox ("Show 'refine_disparity_supertiles'", this.show_refine_supertiles); Loading Loading @@ -3264,6 +3271,8 @@ public class EyesisCorrectionParameters { this.plSnapDispWeight= gd.getNextNumber(); this.plSnapZeroMode= (int) gd.getNextNumber(); this.dbg_migrate= gd.getNextBoolean(); this.show_ortho_combine= gd.getNextBoolean(); this.show_refine_supertiles=gd.getNextBoolean(); this.show_bgnd_nonbgnd= gd.getNextBoolean(); // first on second pass Loading
src/main/java/MeasuredLayers.java 0 → 100644 +513 −0 Original line number Diff line number Diff line /** ** MeasuredLayer - per-tile measured disparity/strength pairs, ** multiple layers can be used for 4-disparity and 2 (hor/vert) pairs ** separately ** ** Copyright (C) 2017 Elphel, Inc. ** ** -----------------------------------------------------------------------------** ** ** MeasuredLayer.java is free software: you can redistribute it and/or modify ** it under the terms of the GNU General Public License as published by ** the Free Software Foundation, either version 3 of the License, or ** (at your option) any later version. ** ** This program is distributed in the hope that it will be useful, ** but WITHOUT ANY WARRANTY; without even the implied warranty of ** MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ** GNU General Public License for more details. ** ** You should have received a copy of the GNU General Public License ** along with this program. If not, see <http://www.gnu.org/licenses/>. ** -----------------------------------------------------------------------------** ** */ public class MeasuredLayers { private DispStrength [][] layers; private int tilesX; private int tilesY; private int superTileSize; private double [][] lapWeight = null; public class DispStrength { public double disparity = Double.NaN; public double strength = 0.0; public DispStrength ( double disparity, double strength) { this.disparity = disparity; this.strength = strength; if (Double.isNaN(disparity)){ this.strength = 0.0; } } public double [] getDispStrength() { double [] ds = {disparity, strength}; return ds; } public void setDisparity(double disparity) { this.disparity = disparity; if (Double.isNaN(disparity)){ this.strength = 0.0; } } public void setStrength(double strength) { this.strength = strength; } public double getDisparity() { return disparity; } public double getStrength() { return strength; } } /** * Create a set of measured layers. Several layers can be used to represent alternative measurements * (with different preset disparity) or for different modes (4 pair correlation, vertical, horizontal) * @param num_layers number of alternative layers used (3 for quad, hor, vert) * @param tilesX number of tiles horizontally in the image * @param tilesY number of tiles vertically in the image * @param superTileSize size of the supertile square */ public MeasuredLayers( int num_layers, int tilesX, int tilesY, int superTileSize) { layers = new DispStrength [num_layers][]; this.tilesX = tilesX; this.tilesY = tilesY; this.superTileSize = superTileSize; this.lapWeight = getLapWeights(); } /** * Get number of tiles in the image horizontally * @return tilesX */ public int getTilesX() { return tilesX; } /** * Get number of tiles in the image vertically * @return tilesY */ public int getTilesY() { return tilesY; } /** * Get number of tiles in each supertile in eac direction * @return superTileSize */ public int getSuperTileSize() { return superTileSize; } /** * Set measured layer * @param num_layer number of the layer to set * @param disparity array of per-tile disparity values (linescan order), * NaN deletes the tile * @param strength array of per-tile disparity values (linescan order), * 0.0 strength is OK - does not delete the tile * @param selection optional tile selection (or null). If unselected, tile is deleted */ public void setLayer ( int num_layer, double [] disparity, double [] strength, boolean [] selection) // may be null { if (layers[num_layer] == null) { int llen = disparity.length; if (layers[num_layer] == null) { layers[num_layer] = new DispStrength [llen]; } } for (int i = 0; i < disparity.length; i++){ if ((selection != null) && !selection[i]){ layers[num_layer][i] = null; } else { if (Double.isNaN(disparity[i])){ layers[num_layer][i] = null; } else { if (layers[num_layer][i] != null){ layers[num_layer][i].setDisparity(disparity[i]); layers[num_layer][i].setStrength(strength[i]); } else { layers[num_layer][i] = new DispStrength(disparity[i], strength[i]); } } } } } /** * Set disparity values for selected measurement layer * @param num_layer number of the layer to set * @param disparity array of per-tile disparity values (linescan order), * NaN deletes the tile. New tile set strength = 0.0 */ public void setDisparity ( int num_layer, double [] disparity) { if (layers[num_layer] == null) { int llen = disparity.length; if (layers[num_layer] == null) { layers[num_layer] = new DispStrength [llen]; } } for (int i = 0; i < disparity.length; i++){ if (Double.isNaN(disparity[i])){ layers[num_layer][i] = null; } else { if (layers[num_layer][i] != null){ layers[num_layer][i].setDisparity(disparity[i]); } else { layers[num_layer][i] = new DispStrength(disparity[i], 0.0); } } } } /** * Set strength values for selected measurement layer. Does not allocate new tiles * if that element was null * @param num_layer number of the layer to set * @param strength array of per-tile strength values (linescan order) */ public void setStrength ( int num_layer, double [] strength) { if (layers[num_layer] != null) { for (int i = 0; i < strength.length; i++){ if (layers[num_layer][i] != null){ layers[num_layer][i].setStrength(strength[i]); } } } } /** * Change tile selection for the layer. Can only unselect existing tiles, not add new ones * @param num_layer number of the layer to set * @param selection array of per-tile boolean selection values (linescan order), */ public void setSelection ( int num_layer, boolean [] selection) { for (int i = 0; i < selection.length; i++){ if (!selection[i]){ layers[num_layer][i] = null; } } } /** * Get total number of the measurement layers in this object * @return number of layers (some may be uninitialized */ public int getnumLayers() { if (layers == null){ return 0; } else { return layers.length; } } /** * Get array of disparity values for the selected layer * @param num_layer number of the layer to read * @return array of per-tile disparity values (in linescan order), * Double.NaN for missing tiles */ public double [] getDisparity( int num_layer) { double [] disparity = new double [layers[num_layer].length]; for (int i = 0; i < disparity.length; i++){ if (layers[num_layer][i] == null){ disparity[i] = Double.NaN; } else { disparity[i] = layers[num_layer][i].getDisparity(); } } return disparity; } /** * Get array of correlation strength values for the selected layer * @param num_layer number of the layer to read * @return array of per-tile strength values (in linescan order), * 0.0 for missing tiles */ public double [] getStrength( int num_layer) { double [] strength = new double [layers[num_layer].length]; for (int i = 0; i < strength.length; i++){ if (layers[num_layer][i] == null){ strength[i] = 0.0; } else { strength[i] = layers[num_layer][i].getStrength(); } } return strength; } /** * Get array of existing tiles for the selected layer * @param num_layer number of the layer to read * @return boolean (per tile) array of existing tiles */ public boolean [] getSelection( int num_layer) { boolean [] selection = new boolean [layers[num_layer].length]; for (int i = 0; i < selection.length; i++){ selection[i] = layers[num_layer][i] != null; } return selection; } /** * Calculate weights for overlapping supertiles to multiply correlation strengths * @return square 2-d array of weights 0.0 ... 1.0, each dimension twice the * supertile size */ private double [][] getLapWeights(){ final double [][] lapWeight = new double [2 * superTileSize][2 * superTileSize]; final double [] lapWeight1d = new double [superTileSize]; final int superTileSize2 = 2 * superTileSize; for (int i = 0; i < superTileSize; i++){ lapWeight1d[i] = 0.5*(1.0 - Math.cos((i + 0.5)* Math.PI/superTileSize)); } for (int i = 0; i < superTileSize; i++){ for (int j = 0; j < superTileSize; j++){ lapWeight[i] [ j] = lapWeight1d[i]*lapWeight1d[j]; lapWeight[superTileSize2 - 1 - i][ j] = lapWeight[i][j]; lapWeight[i] [superTileSize2 - 1 - j] = lapWeight[i][j]; lapWeight[superTileSize2 - 1 - i][superTileSize2 - 1 - j] = lapWeight[i][j]; } } double s = 0.0; for (int i = 0; i < superTileSize2; i++){ for (int j = 0; j < superTileSize2; j++){ s+=lapWeight[i][j]; } } System.out.println("getLapWeights: sum = "+s); return lapWeight; } /** * Get selection for the specific measurement layer and supertile X,Y coordinates * in the image. Combined with input selection * @param num_layer number of the measurement layer to process * @param stX supertile horizontal position in the image * @param stY supertile vertical position in the image * @param sel_in optional selection for this supertile (linescan, 4 * supetile size) * @param null_if_none return null if there are no selected tiles in teh result selection * @return boolean array [4*superTileSize] of the selected tiles on the specified * measurement layer */ public boolean [] getSupertileSelection( int num_layer, int stX, int stY, boolean [] sel_in, double strength_floor, boolean null_if_none) { if ((layers[num_layer] == null) && null_if_none){ return null; } int st2 = 2 * superTileSize; int st_half = superTileSize/2; boolean [] selection = new boolean [st2 * st2]; int num_selected = 0; if (layers[num_layer] != null) { for (int dy = 0; dy < st2; dy ++){ int y = superTileSize * stY -st_half + dy; if ((y >= 0) && (y < tilesY)) { for (int dx = 0; dx < st2; dx ++){ int x = superTileSize * stX -st_half + dx; if ((x >= 0) && (x < tilesX)) { int indx = y * tilesX + x; int indx_st = dy * st2 + dx; if (((sel_in == null) || sel_in[indx_st]) && (layers[num_layer][indx] != null) && (layers[num_layer][indx].getStrength() >= strength_floor)){ selection[indx_st] = true; num_selected ++; } } } } } } if (null_if_none && (num_selected == 0)) return null; return selection; } /** * Get selection for the specific measurement layer and supertile X,Y coordinates * in the image and combine with disparity far/near limits. Alse combined with * input selection * @param num_layer number of the measurement layer to process * @param stX supertile horizontal position in the image * @param stY supertile vertical position in the image * @param sel_in optional selection for this supertile (linescan, 4 * supetile size) * @param disp_far lowest acceptable disparity value. Double.NaN - do not check. * @param disp_near highest acceptable disparity value. Double.NaN - do not check. * @param null_if_none return null if there are no selected tiles in teh result selection * @return boolean array [4*superTileSize] of the selected tiles on the specified * measurement layer */ public boolean [] getSupertileSelection( int num_layer, int stX, int stY, boolean [] sel_in, double disp_far, double disp_near, double strength_floor, boolean null_if_none) { if ((layers[num_layer] == null) && null_if_none){ return null; } int st2 = 2 * superTileSize; int st_half = superTileSize/2; boolean [] selection = new boolean [st2 * st2]; int num_selected = 0; if (layers[num_layer] != null) { for (int dy = 0; dy < st2; dy ++){ int y = superTileSize * stY -st_half + dy; if ((y >= 0) && (y < tilesY)) { for (int dx = 0; dx < st2; dx ++){ int x = superTileSize * stX -st_half + dx; if ((x >= 0) && (x < tilesX)) { int indx = y * tilesX + x; int indx_st = dy * st2 + dx; if (((sel_in == null) || sel_in[indx_st]) && (layers[num_layer][indx] != null)){ if ( (Double.isNaN(disp_far) || (layers[num_layer][indx].getDisparity() >= disp_far)) && (Double.isNaN(disp_near) || (layers[num_layer][indx].getDisparity() <= disp_near)) && (layers[num_layer][indx].getStrength() >= strength_floor)){ selection[indx_st] = true; num_selected ++; } } } } } } } if (null_if_none && (num_selected == 0)) return null; return selection; } /** * Get number of "true" elements in a boolean array. Null is OK, it results in 0 * @param selected boolean array to count set elements in * @return number of selected elements */ public static int getNumSelected( boolean [] selected) { if (selected == null) return 0; int num_selected = 0; for (int i = 0; i < selected.length; i++){ if (selected[i]) num_selected ++; } return num_selected; } public static double getSumStrength( double [][] disp_strength) { if (disp_strength == null) return 0.0; double sw = 0.0; for (int i = 0; i < disp_strength[1].length; i++){ sw += disp_strength[1][i]; } return sw; } /** * Get disparity and correlation strength for the specific measurement layer and * supertile X,Y coordinates in the image. Combined with input selection * @param num_layer number of the measurement layer to process * @param stX supertile horizontal position in the image * @param stY supertile vertical position in the image * @param sel_in optional selection for this supertile (linescan, 4 * supetile size) * @param null_if_none return null if there are no selected tiles in the result selection * @param strength_pow Non-linear treatment of correlation strength. Raise data to the strength_pow power * @return double [2][4*superTileSize] {disparity[4*superTileSize], strength [4*superTileSize]} */ public double[][] getDisparityStrength ( int num_layer, int stX, int stY, boolean [] sel_in, double strength_floor, double strength_pow, boolean null_if_none) { if ((layers[num_layer] == null) && null_if_none){ return null; } int st2 = 2 * superTileSize; int st_half = superTileSize/2; double [][] ds = new double [2][st2*st2]; int num_selected = 0; if (layers[num_layer] != null) { for (int dy = 0; dy < st2; dy ++){ int y = superTileSize * stY -st_half + dy; if ((y >= 0) && (y < tilesY)) { for (int dx = 0; dx < st2; dx ++){ int x = superTileSize * stX -st_half + dx; if ((x >= 0) && (x < tilesX)) { int indx = y * tilesX + x; int indx_st = dy * st2 + dx; if (((sel_in == null) || sel_in[indx_st]) && (layers[num_layer][indx] != null)){ ds[0][indx_st] = layers[num_layer][indx].getDisparity(); double w = layers[num_layer][indx].getStrength() - strength_floor; if (w > 0) { if (strength_pow != 1.0) w = Math.pow(w, strength_pow); w *= lapWeight[dy][dx]; ds[0][indx_st] = layers[num_layer][indx].getDisparity(); ds[1][indx_st] = w; num_selected ++; } } } } } } } if (null_if_none && (num_selected == 0)) return null; return ds; } }
