Developing dual-max LMA. Before simultaneous fitting

src/main/java/com/elphel/imagej/tileprocessor/Corr2dLMA.java

@@ -547,7 +547,6 @@ public class Corr2dLMA {
 			}
 		}
 		
-		
 		// apply corrections
 		if (apply) {
 			for (int ncam = 0; ncam <num_cams; ncam++) {
@@ -1636,7 +1635,28 @@ public class Corr2dLMA {
 		return xy_offsets;
 	}
 	
-	
+	/**
+	 * Calculate offsets for multiple disparities, for each defined pair x,y expected offset, assuming only disparity, not lazy eye   
+	 * @param corrs - pairs 2D correlations (each in scanline order) - just to determine null/non-null
+	 * @param disparity_strengths - expected {disparity, strength} pairs (e.g. from CM)
+	 * @param disp_dist - per camera disparity matrix as a 1d (linescan order))
+	 * @return array of per-disparity, per pair x,y expected center offset in 2D correlations or nulls for undefined pairs (or null if already set)
+	 */
+	public double [][][] getPairsOffsets(
+			double [][]       corrs,
+			boolean []        pair_mask,
+			double [][]       disparity_strengths,
+			double [][]       disp_dist){ //
+		double [][][] pairs_offsets = new double [disparity_strengths.length][][];
+		for (int i = 0; i < pairs_offsets.length; i++) {
+			pairs_offsets[i] = getPairsOffsets(
+					corrs,                     // double [][]       corrs,
+					pair_mask,                 // boolean []        pair_mask,
+					disparity_strengths[i][0], // double            disparity,
+					disp_dist);                // double [][]       disp_dist)
+		}
+		return pairs_offsets;
+	}
 
 	public void printParams() { // not used in lwir
 		// to make sure it is updated
@@ -2062,7 +2082,7 @@ public class Corr2dLMA {
 		double [][] maxmin_amp = getMaxMinAmpTile();
 		double [][] abc =        getABCTile();
 		for (int tile = 0; tile < numTiles; tile++) {
-			ds[tile][0] = Double.NaN;
+			ds[tile][0] = Double.NaN + 0;
 			if (Double.isNaN(maxmin_amp[tile][0])) {
 				continue;
 			}

src/main/java/com/elphel/imagej/tileprocessor/ImageDtt.java

@@ -2631,7 +2631,8 @@ public class ImageDtt extends ImageDttCPU {
 								if (debugTile1) {
 									correlation2d.corrLMA2DualMax( // null pointer
 											imgdtt_params,                // ImageDttParameters  imgdtt_params,
-											imgdtt_params.lmas_LY_single, // false,    // boolean             adjust_ly, // adjust Lazy Eye
+								        	1,                            // int     combine_mode,   // 0 - both,  1 - strongest, 2 - nearest to zero, 3 - FG, 4 - BG
+//											imgdtt_params.lmas_LY_single, // false,    // boolean             adjust_ly, // adjust Lazy Eye
 											corr_wnd,                     // double [][]         corr_wnd, // correlation window to save on re-calculation of the window
 											corr_wnd_inv_limited,         // corr_wnd_limited, // correlation window, limited not to be smaller than threshold - used for finding max/convex areas (or null)
 											corrs,                        // corrs,          // double [][]         corrs,
@@ -2644,7 +2645,6 @@ public class ImageDtt extends ImageDttCPU {
 											imgdtt_params.ortho_vasw_pwr, // double    vasw_pwr,  // value as weight to this power,
 											null, // debug_lma_tile,               // double []           debug_lma_tile,
 											(debugTile0 ? 1: -2),         // int                 debug_level,
-//											-2, //0,                            // tile_lma_debug_level, // +2,         // int                 debug_level,
 											tileX,                        // int                 tileX, // just for debug output
 											tileY ); 
 									

src/main/java/com/elphel/imagej/tileprocessor/ImageDttParameters.java

@@ -77,6 +77,14 @@ public class ImageDttParameters {
 	public int     dbg_pair_mask =          0x3f;  // which pairs to combine
 	public int     corr_strip_hight =       9;     // number of rows to calculate
 
+	//bimax_
+	public double  bimax_lpf_neib =          0.5; // if >0, add ortho neibs (corners - squared)
+	public double  bimax_notch_pwr =         4.0; // calculating notch filter to suppress input data of the weaker maximum 
+	public double  bimax_adv_power =         2.0; // reduce weight from overlap with adversarial maximum 
+	public int     bimax_rad_convex_search = 2;   // how far from predicted to search for maximums
+	public int     bimax_min_num_samples =   4;   // minimal number of samples per pair per maximum
+	public int     bimax_min_num_pairs =     8;   // minimal number of used pairs
+
 	//lmamask_
 	public boolean lmamask_dbg =              false;  // show LMA images, exit after single BG
 	public boolean lmamask_en =               false;  // Use disparity-based LMA samples filter
@@ -450,6 +458,21 @@ public class ImageDttParameters {
 			gd.addNumericField("Number of correlation rows to combine (strip height)",            this.corr_strip_hight,  0, 3, "",
 					"Number of rows to combine/interpolate correlation results. Rows are twice denser than pixels correponding to largest baseline disparity");
 
+			gd.addMessage("LMA prefiltering for dual disparities in a tile");
+			
+		    gd.addNumericField("LPF for combined maximums",                                       this.bimax_lpf_neib,  6,8,"",
+		    		"0 - no LPF, >0 - add 4 scaled ortho neighbors and 4 corners bscaled as square of this value");
+		    gd.addNumericField("Notch filter power",                                              this.bimax_notch_pwr,  6,8,"",
+		    		"Reduce offending (other) correlation maximum by multiplying. High power reduces all < 1.0 values");
+		    gd.addNumericField("Reduction of sample weight by the other maximum",                 this.bimax_adv_power,  6,8,"",
+		    		"If X= (this strength /other strength), scale by X/(1+X) to theis power");
+			gd.addNumericField("Convex max search around predicted disparity",                    this.bimax_rad_convex_search,  0, 3, "",
+					"How far from predicted to search for maximums");
+			gd.addNumericField("Minimal samples per pair per max",                                this.bimax_min_num_samples,  0, 3, "",
+					"Minimal number of samples per pair per maximum to use this pair by LMA (each of the maximums used)");
+			gd.addNumericField("Minimal number of used pairs",                                    this.bimax_min_num_pairs,  0, 3, "",
+					"Do not use LMA if total number of used pairs is lower");
+			
 			gd.addMessage("LMA samples filter based on estimated disparity");
 			gd.addCheckbox    ("Debug LMA",                                                       this.lmamask_dbg,
 					"Generate debug images and exit after first clt_process_tl_correlations() while generating background image");
@@ -859,6 +882,13 @@ public class ImageDttParameters {
   			this.dbg_pair_mask=           (int) gd.getNextNumber();
   			this.corr_strip_hight=        (int) gd.getNextNumber();
   			
+  		    this.bimax_lpf_neib =               gd.getNextNumber();
+  		    this.bimax_notch_pwr =              gd.getNextNumber();
+  		    this.bimax_adv_power =              gd.getNextNumber();
+  		    this.bimax_rad_convex_search= (int) gd.getNextNumber();
+  		    this.bimax_min_num_samples=   (int) gd.getNextNumber();
+  		    this.bimax_min_num_pairs=     (int) gd.getNextNumber();
+  			
   			this.lmamask_dbg =              gd.getNextBoolean();
   			this.lmamask_en =               gd.getNextBoolean();
 			this.lmamask_magic =            gd.getNextNumber();
@@ -1078,6 +1108,13 @@ public class ImageDttParameters {
 		properties.setProperty(prefix+"dbg_pair_mask",        this.dbg_pair_mask +"");
 		properties.setProperty(prefix+"corr_strip_hight",     this.corr_strip_hight +"");
 		
+		properties.setProperty(prefix+"bimax_lpf_neib",          this.bimax_lpf_neib +"");
+		properties.setProperty(prefix+"bimax_notch_pwr",         this.bimax_notch_pwr +"");
+		properties.setProperty(prefix+"bimax_adv_power",         this.bimax_adv_power +"");
+		properties.setProperty(prefix+"bimax_rad_convex_search", this.bimax_rad_convex_search +"");
+		properties.setProperty(prefix+"bimax_min_num_samples",   this.bimax_min_num_samples +"");
+		properties.setProperty(prefix+"bimax_min_num_pairs",     this.bimax_min_num_pairs +"");
+
 		properties.setProperty(prefix+"lmamask_dbg",          this.lmamask_dbg +"");
 		properties.setProperty(prefix+"lmamask_en",           this.lmamask_en +"");
 		properties.setProperty(prefix+"lmamask_magic",        this.lmamask_magic +"");
@@ -1301,6 +1338,12 @@ public class ImageDttParameters {
 		if (properties.getProperty(prefix+"dbg_pair_mask")!=null)        this.dbg_pair_mask=Integer.parseInt(properties.getProperty(prefix+"dbg_pair_mask"));
 		if (properties.getProperty(prefix+"corr_strip_hight")!=null)     this.corr_strip_hight=Integer.parseInt(properties.getProperty(prefix+"corr_strip_hight"));
 
+		if (properties.getProperty(prefix+"bimax_lpf_neib")!=null)          this.bimax_lpf_neib=Double.parseDouble(properties.getProperty(prefix+"bimax_lpf_neib"));
+		if (properties.getProperty(prefix+"bimax_notch_pwr")!=null)         this.bimax_notch_pwr=Double.parseDouble(properties.getProperty(prefix+"bimax_notch_pwr"));
+		if (properties.getProperty(prefix+"bimax_adv_power")!=null)         this.bimax_adv_power=Double.parseDouble(properties.getProperty(prefix+"bimax_adv_power"));
+		if (properties.getProperty(prefix+"bimax_rad_convex_search")!=null) this.bimax_rad_convex_search=Integer.parseInt(properties.getProperty(prefix+"bimax_rad_convex_search"));
+		if (properties.getProperty(prefix+"bimax_min_num_samples")!=null)   this.bimax_min_num_samples=Integer.parseInt(properties.getProperty(prefix+"bimax_min_num_samples"));
+		if (properties.getProperty(prefix+"bimax_min_num_pairs")!=null)     this.bimax_min_num_pairs=Integer.parseInt(properties.getProperty(prefix+"bimax_min_num_pairs"));
 		
 		if (properties.getProperty(prefix+"lmamask_dbg")!=null)              this.lmamask_dbg=Boolean.parseBoolean(properties.getProperty(prefix+"lmamask_dbg"));
 		if (properties.getProperty(prefix+"lmamask_en")!=null)               this.lmamask_en=Boolean.parseBoolean(properties.getProperty(prefix+"lmamask_en"));
@@ -1542,6 +1585,13 @@ public class ImageDttParameters {
 		idp.dbg_pair_mask=           this.dbg_pair_mask;
 		idp.corr_strip_hight=        this.corr_strip_hight;
 
+		idp.bimax_lpf_neib=          this.bimax_lpf_neib;
+		idp.bimax_notch_pwr=         this.bimax_notch_pwr;
+		idp.bimax_adv_power=         this.bimax_adv_power;
+		idp.bimax_rad_convex_search= this.bimax_rad_convex_search;
+		idp.bimax_min_num_samples=   this.bimax_min_num_samples;
+		idp.bimax_min_num_pairs=     this.bimax_min_num_pairs;
+		
 		idp.lmamask_dbg=               this.lmamask_dbg;
 		idp.lmamask_en=                this.lmamask_en;
 		idp.lmamask_magic=             this.lmamask_magic;

src/main/java/com/elphel/imagej/tileprocessor/OpticalFlow.java

@@ -4361,7 +4361,9 @@ public class OpticalFlow {
 		// Save pair selection and minimize them for scanning, then restore;
 		int num_sensors =scenes[indx_ref].getNumSensors();
 		int save_pairs_selection = clt_parameters.img_dtt.getMcorr(num_sensors);
-		clt_parameters.img_dtt.setMcorr(num_sensors, 0 ); // remove all
+		// FIXME: uncomment next
+		System.out.println ("++++ Uncomment next line ++++");
+////		clt_parameters.img_dtt.setMcorr(num_sensors, 0 ); // remove all
 		clt_parameters.img_dtt.setMcorrNeib(num_sensors,true);
 		clt_parameters.img_dtt.setMcorrSq  (num_sensors,true); // remove even more?
 		clt_parameters.img_dtt.setMcorrDia (num_sensors,true); // remove even more?