Skip to content

I
imagej-elphel
Commit 1dbe6bd1 authored by Andrey Filippov's avatar Andrey Filippov
Browse files
Options

more bugs fixed

parent d2c9aab5
Hide whitespace changes
Inline Side-by-side
Showing with 256 additions and 23 deletions
src/main/java/FocusingField.java
View file @ 1dbe6bd1
......@@ -62,7 +62,12 @@ public class FocusingField {
double filterInputDiff; // um
boolean filterInputFirstLast;
boolean filterInputTooFar; // filter samples that are too far from the "center of mass" of other samples
double filterInpuFarRatio; // remove samples that are farther than this ration of average distance
double filterInputFarRatio; // remove samples that are farther than this ration of average distance
boolean filterInputConcave;
double filterInputConcaveSigma;
boolean filterInputConcaveRemoveFew;
int filterInputConcaveMinSeries;
double filterInputConcaveScale;
// when false - tangential is master
double [] minMeas; // pixels
double [] maxMeas; // pixels
......@@ -172,13 +177,19 @@ public class FocusingField {
filterInputDiff = 2.0; // um
filterInputFirstLast = true;
filterInputTooFar = true; // filter samples that are too far from the "center of mass" of other samples
filterInpuFarRatio = 3.0; // remove samples that are farther than this ration of average distance
filterInputFarRatio = 3.0; // remove samples that are farther than this ration of average distance
filterInputConcave = true; //um
filterInputConcaveSigma = 8.0; //um
filterInputConcaveRemoveFew=true;
filterInputConcaveMinSeries=5;
filterInputConcaveScale=0.8;
// when false - tangential is master
double [] minMeasDflt= {0.5,0.5,0.5,0.5,0.5,0.5}; // pixels
minMeas= minMeasDflt; // pixels
double [] maxMeasDflt= {4.5,4.5,4.5,4.5,4.5,4.5}; // pixels
maxMeas= maxMeasDflt; // pixels
double [] thresholdMaxDflt= {2.4,3.0,2.6,3.0,3.1,3.0}; // pixels
// double [] thresholdMaxDflt= {2.4,3.0,2.6,3.0,3.1,3.0}; // pixels
double [] thresholdMaxDflt= {3.5,3.5,3.5,3.5,3.5,3.5}; // pixels
thresholdMax= thresholdMaxDflt; // pixels
useMinMeas= true;
useMaxMeas= true;
......@@ -236,9 +247,9 @@ public class FocusingField {
updateWeightWhileFitting=false;
debugPoint=-1;
debugParameter=-1;
currentPX0=pX0_distortions;
currentPY0=pY0_distortions;
}
......@@ -262,7 +273,15 @@ public class FocusingField {
properties.setProperty(prefix+"filterInputDiff",filterInputDiff+"");
properties.setProperty(prefix+"filterInputFirstLast",filterInputFirstLast+"");
properties.setProperty(prefix+"filterInputTooFar",filterInputTooFar+"");
properties.setProperty(prefix+"filterInpuFarRatio",filterInpuFarRatio+"");
properties.setProperty(prefix+"filterInputFarRatio",filterInputFarRatio+"");
properties.setProperty(prefix+"filterInputConcave",filterInputConcave+"");
properties.setProperty(prefix+"filterInputConcaveSigma",filterInputConcaveSigma+"");
properties.setProperty(prefix+"filterInputConcaveRemoveFew",filterInputConcaveRemoveFew+"");
properties.setProperty(prefix+"filterInputConcaveMinSeries",filterInputConcaveMinSeries+"");
properties.setProperty(prefix+"filterInputConcaveScale",filterInputConcaveScale+"");
for (int chn=0; chn<minMeas.length; chn++) properties.setProperty(prefix+"minMeas_"+chn,minMeas[chn]+"");
for (int chn=0; chn<maxMeas.length; chn++) properties.setProperty(prefix+"maxMeas_"+chn,maxMeas[chn]+"");
for (int chn=0; chn<thresholdMax.length; chn++) properties.setProperty(prefix+"thresholdMax_"+chn,thresholdMax[chn]+"");
......@@ -323,8 +342,22 @@ public class FocusingField {
filterInputFirstLast=Boolean.parseBoolean(properties.getProperty(prefix+"filterInputFirstLast"));
if (properties.getProperty(prefix+"filterInputTooFar")!=null)
filterInputTooFar=Boolean.parseBoolean(properties.getProperty(prefix+"filterInputTooFar"));
if (properties.getProperty(prefix+"filterInpuFarRatio")!=null)
filterInpuFarRatio=Double.parseDouble(properties.getProperty(prefix+"filterInpuFarRatio"));
if (properties.getProperty(prefix+"filterInputFarRatio")!=null)
filterInputFarRatio=Double.parseDouble(properties.getProperty(prefix+"filterInputFarRatio"));
if (properties.getProperty(prefix+"filterInputConcave")!=null)
filterInputConcave=Boolean.parseBoolean(properties.getProperty(prefix+"filterInputConcave"));
if (properties.getProperty(prefix+"filterInputConcaveSigma")!=null)
filterInputConcaveSigma=Double.parseDouble(properties.getProperty(prefix+"filterInputConcaveSigma"));
if (properties.getProperty(prefix+"filterInputConcaveRemoveFew")!=null)
filterInputConcaveRemoveFew=Boolean.parseBoolean(properties.getProperty(prefix+"filterInputConcaveRemoveFew"));
if (properties.getProperty(prefix+"filterInputConcaveMinSeries")!=null)
filterInputConcaveMinSeries=Integer.parseInt(properties.getProperty(prefix+"filterInputConcaveMinSeries"));
if (properties.getProperty(prefix+"filterInputConcaveScale")!=null)
filterInputConcaveScale=Double.parseDouble(properties.getProperty(prefix+"filterInputConcaveScale"));
for (int chn=0; chn<minMeas.length; chn++) if (properties.getProperty(prefix+"minMeas_"+chn)!=null)
minMeas[chn]=Double.parseDouble(properties.getProperty(prefix+"minMeas_"+chn));
for (int chn=0; chn<maxMeas.length; chn++) if (properties.getProperty(prefix+"maxMeas_"+chn)!=null)
......@@ -473,11 +506,17 @@ public boolean configureDataVector(String title, boolean forcenew, boolean enabl
gd.addNumericField("Maximal allowed PSF FWHM variations fro the move above",filterInputDiff,3,5,"um");
gd.addCheckbox("Remove first/last in a series of measuremnts separated by small (see above) steps",filterInputFirstLast);
gd.addCheckbox("Remove measurements taken too far from the rest for the same channel/sample",filterInputTooFar);
gd.addNumericField("\"Too far\" ratio to the average distance to the center of measurements",filterInpuFarRatio,3,5,"um");
gd.addNumericField("\"Too far\" ratio to the average distance to the center of measurements",filterInputFarRatio,3,5,"um");
gd.addCheckbox("Filter non-concave areas from best focus for each sample",filterInputConcave);
gd.addNumericField("Concave filter sigma",filterInputConcaveSigma,3,5,"um");
gd.addCheckbox("Remove small series ",filterInputConcaveRemoveFew);
gd.addNumericField("Minimal number of samples (to remove / apply concave vilter) ",filterInputConcaveMinSeries,3,5,"samples");
gd.addNumericField("Concave filter sigma",filterInputConcaveScale,3,5,"<=1.0");
gd.addCheckbox("Sagittal channels are master channels (false - tangential are masters)",sagittalMaster);
gd.addMessage("=== Setting minimal measured PSF radius for different colors/directions ===");
for (int i=0;i<minMeas.length;i++){
gd.addNumericField(tmpFieldFitting.getDescription(i),this.minMeas[i],3,5,"pix");
......@@ -519,14 +558,19 @@ public boolean configureDataVector(String title, boolean forcenew, boolean enabl
// boolean configureDataVector(String title, boolean forcenew, boolean moreset)
parallelOnly= gd.getNextBoolean();
filterInput= gd.getNextBoolean();
filterInputMotorDiff= gd.getNextNumber();
filterInputDiff= gd.getNextNumber();
parallelOnly= gd.getNextBoolean();
filterInput= gd.getNextBoolean();
filterInputMotorDiff= gd.getNextNumber();
filterInputDiff= gd.getNextNumber();
filterInputFirstLast= gd.getNextBoolean();
filterInputTooFar= gd.getNextBoolean();
filterInpuFarRatio= gd.getNextNumber();
filterInputFirstLast= gd.getNextBoolean();
filterInputTooFar= gd.getNextBoolean();
filterInputFarRatio= gd.getNextNumber();
filterInputConcave= gd.getNextBoolean();
filterInputConcaveSigma= gd.getNextNumber();
filterInputConcaveRemoveFew= gd.getNextBoolean();
filterInputConcaveMinSeries= (int) gd.getNextNumber();
filterInputConcaveScale= gd.getNextNumber();
sagittalMaster= gd.getNextBoolean();
for (int i=0;i<minMeas.length;i++)this.minMeas[i]= gd.getNextNumber();
......@@ -586,6 +630,180 @@ private void maskDataWeights(boolean [] enable){
}
}
private boolean [] filterConcave(
double sigma,
boolean removeInsufficient,
int minSeries,
double concaveScale,
boolean [] enable_in){
if (enable_in==null) {
enable_in=new boolean [dataVector.length];
for (int i=0;i<enable_in.length;i++)enable_in[i]=true;
}
// int minPoints=minSeries;
int debugThreshold=1;
double maxGap=sigma; // this point has this gap towards minimal
double kexp=-0.5/(sigma*sigma);
boolean [] enable_out=enable_in.clone();
double keepNearMin=sigma; // when removing non-concave points around min, skip very close ones
double [][] flatSampleCoordinates=fieldFitting.getSampleCoordinates();
int numFilteredInsufficient = 0;
int numFiltered = 0;
int [][] numPoints=new int [getNumChannels()][getNumSamples()];
for (int chn=0;chn<numPoints.length;chn++) for (int sample=0;sample<numPoints[chn].length;sample++) numPoints[chn][sample]=0;
for (int index=0;index<dataVector.length;index++) if ((index>=enable_in.length) ||enable_in[index]){
numPoints[dataVector[index].channel][dataVector[index].sampleIndex]++;
}
int [][][] indices=new int[numPoints.length][numPoints[0].length][];
for (int chn=0;chn<numPoints.length;chn++) for (int sample=0;sample<numPoints[chn].length;sample++){
indices[chn][sample]=new int [numPoints[chn][sample]]; // may be 0 length
numPoints[chn][sample]=0; // will be used as a counter
}
for (int index=0;index<dataVector.length;index++) if ((index>=enable_in.length) ||enable_in[index]){
int chn=dataVector[index].channel;
int sample=dataVector[index].sampleIndex;
// numPoints[dataVector[index].channel][dataVector[index].sampleIndex]++;
indices[chn][sample][numPoints[chn][sample]++]=index;
}
for (int chn=0;chn<numPoints.length;chn++) for (int sample=0;sample<numPoints[chn].length;sample++){
if (indices[chn][sample].length<minSeries){
if (indices[chn][sample].length>0) {
if (debugLevel>0) System.out.println("filterConcave(): Channel "+chn+" sample "+sample+" has too few points - "+indices[chn][sample].length+" < "+minSeries);
if (removeInsufficient){
for (int i=0;i<indices[chn][sample].length;i++){
enable_out[indices[chn][sample][i]]=false;
numFilteredInsufficient++;
}
}
}
} else {
int [] thisIndices=indices[chn][sample];
double [] point_z = new double[thisIndices.length];
double [] point_v = new double[thisIndices.length];
double [] point_filt = new double[thisIndices.length];
double [] point_vdz = new double[thisIndices.length];
double [] point_slope = new double[thisIndices.length];
boolean [] nonConcave=new boolean[thisIndices.length];
for (int i=0;i<thisIndices.length;i++){
//if ((chn==0) && (sample==7) && (i>=16)){
// System.out.println("DEBUG00");
//}
point_z[i]=fieldFitting.getMotorsZ(
dataVector[thisIndices[i]].motors, // 3 motor coordinates
flatSampleCoordinates[sample][0], // pixel x
flatSampleCoordinates[sample][1]); // pixel y
point_v[i]=dataVector[thisIndices[i]].value;
nonConcave[i]=false;
}
for (int i=0;i<thisIndices.length;i++){
point_filt[i]=0.0;
double weight=0.0;
for (int j=0;j<thisIndices.length;j++){
double r=point_z[i]-point_z[j];
double w=Math.exp(kexp*r*r);
weight+=w;
point_filt[i]+=w*point_v[j];
}
point_filt[i]/=weight;
}
for (int i=0;i<thisIndices.length;i++){
point_vdz[i]=0.0;
double S0=0.0,SX=0.0,SY=0.0,SX2=0.0,SXY=0.0;
for (int j=0;j<thisIndices.length;j++){
double x=point_z[j]-point_z[i];
double v=point_filt[j];
double w=Math.exp(kexp*x*x);
S0+=w;
SX+=w*x;
SY+=w*v;
SX2+=w*x*x;
SXY+=w*x*v;
}
point_vdz[i]=(SXY*S0-SX*SY)/(SX2*S0-SX*SX);
}
// find min on filtered
int minIndex=0;
for (int i=1;i<thisIndices.length;i++) if (point_filt[i]<point_filt[minIndex]) minIndex=i;
for (int i=0;i<thisIndices.length;i++) {
if (i == minIndex) point_slope[i]=0.0;
else point_slope[i]= (point_filt[i]-point_filt[minIndex])/(point_z[i]-point_z[minIndex]);
}
//concaveScale
for (int i=0;i<thisIndices.length;i++) {
if (
(((point_z[i]-point_z[minIndex])>keepNearMin) && (concaveScale*point_slope[i]>point_vdz[i])) ||
(((point_z[minIndex]-point_z[i])>keepNearMin) && (concaveScale*point_slope[i]<point_vdz[i]))){
nonConcave[i]=true;
}
}
// find gaps double maxGap=sigma; // this point has this gap towards minimal
for (int i=0;i<thisIndices.length;i++) if (!nonConcave[i]){
if ((point_z[i]-point_z[minIndex])>keepNearMin){
boolean goodPoint=false;
for (int j=0;j<thisIndices.length;j++) if (
(point_z[j]>point_z[minIndex]) &&
(point_z[j]<point_z[i]) &&
((point_z[i]-point_z[j]) < maxGap ) ){
goodPoint=true;
break;
}
if (!goodPoint) nonConcave[i]=true;
} else if ((point_z[minIndex]-point_z[i])>keepNearMin) {
boolean goodPoint=false;
for (int j=0;j<thisIndices.length;j++) if (
(point_z[j]<point_z[minIndex]) &&
(point_z[j]>point_z[i]) &&
((point_z[j]-point_z[i]) < maxGap ) ){
goodPoint=true;
break;
}
if (!goodPoint) nonConcave[i]=true;
}
}
// propagate
for (int i=0;i<thisIndices.length;i++) if (!nonConcave[i]){
if ((point_z[i]-point_z[minIndex])>keepNearMin){
for (int j=0;j<thisIndices.length;j++) if (
nonConcave[j] &&
(point_z[j]>point_z[minIndex]) &&
(point_z[j]<point_z[i])){
nonConcave[i]=true;
break;
}
} else if ((point_z[minIndex]-point_z[i])>keepNearMin) {
for (int j=0;j<thisIndices.length;j++) if (
nonConcave[j] &&
(point_z[j]<point_z[minIndex]) &&
(point_z[j]>point_z[i])){
nonConcave[i]=true;
break;
}
}
}
for (int i=0;i<thisIndices.length;i++) if (nonConcave[i]){
enable_out[thisIndices[i]]=false;
numFiltered++;
}
if (debugLevel>debugThreshold) {
System.out.println("filterConcave(), chn="+chn+", sample="+sample);
for (int i=0;i<thisIndices.length;i++){
System.out.println(i+": z="+ IJ.d2s(point_z[i],3)+", v="+ IJ.d2s(point_v[i],3)+
", filt="+ IJ.d2s(point_filt[i],3)+", vdz="+ IJ.d2s(100*point_vdz[i],3)+
", slope="+ IJ.d2s(100*point_slope[i],3)+ ", concave="+(nonConcave[i]?0.0:1.0));
}
}
}
}
if (debugLevel>0) System.out.println("filterConcave(): removed for too few points "+numFilteredInsufficient+" samples");
if (debugLevel>0) System.out.println("filterConcave(): removed for non-concave "+numFiltered+" samples");
return enable_out;
}
private boolean [] filterTooFar(double ratio,boolean [] enable_in){
if (enable_in==null) {
enable_in=new boolean [dataVector.length];
......@@ -796,11 +1014,23 @@ public void setDataVector(MeasuredSample [] vector){ // remove unused channels i
if (filterInputTooFar){
boolean [] en=dataWeightsToBoolean();
en= filterTooFar(
filterInpuFarRatio,
filterInputFarRatio,
en);
maskDataWeights(en);
}
if (filterInputConcave){
boolean [] en=dataWeightsToBoolean();
en= filterConcave(
filterInputConcaveSigma,
filterInputConcaveRemoveFew,
filterInputConcaveMinSeries,
filterInputConcaveScale,
en);
maskDataWeights(en);
}
}
// for compatibility with Distortions class\
......@@ -2672,8 +2902,8 @@ public boolean LevenbergMarquardt(boolean openDialog, int debugLevel){
this.pX0_distortions=pX0;
this.pY0_distortions=pY0;
// copy distortions to current PX0/PY0
this.currentPX0=pX0_distortions;
this.currentPY0=pY0_distortions;
// this.currentPX0=pX0_distortions;
// this.currentPY0=pY0_distortions;
this.sampleCoord=sampleCoord;
this.measurements=new ArrayList<FocusingFieldMeasurement>();
this.stopRequested=stopRequested;
......@@ -2826,7 +3056,8 @@ public boolean LevenbergMarquardt(boolean openDialog, int debugLevel){
private int [][] sampleCorrChnParIndex=null;
private boolean [] dflt_sampleCorrSelect= {false,false,false,false};
private double [] dflt_sampleCorrCost= {1.0,50.0,1.0,1.0};
// private double [] dflt_sampleCorrCost= {1.0,50.0,1.0,1.0};
private double [] dflt_sampleCorrCost= {0.3,20.0,0.3,1.0};
private double dflt_sampleCorrSigma= 2.0; // mm
private double dflt_sampleCorrPullZero= 0.75; // fraction
public final String [] channelDescriptions={
......@@ -3598,7 +3829,9 @@ if ((chn==4) && (sampleIndex==3)){
}
*/
}
public double [][] getSampleCoordinates(){
return sampleCoordinates;
}
public double [] getCorrPar(int chn, int sampleIndex){
/* if ((sampleCorrChnParIndex==null) || (sampleCorrChnParIndex[chn]==null)) return null;
double [] corr =new double [sampleCorrChnParIndex[chn].length];
......
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