!51 增加多元线性回归可调参数

Merge pull request !51 from 逐光/test

src/main/java/org/wlld/imageRecognition/Convolution.java

@@ -64,13 +64,13 @@ public class Convolution extends Frequency {
     }
 
     public List<List<Double>> kAvg(ThreeChannelMatrix threeMatrix, int sqNub
-            , int regionSize) throws Exception {
+            , int regionSize, TempleConfig templeConfig) throws Exception {
         RGBSort rgbSort = new RGBSort();
         List<List<Double>> features = new ArrayList<>();
         List<ThreeChannelMatrix> threeChannelMatrixList = regionThreeChannelMatrix(threeMatrix, regionSize);
         for (ThreeChannelMatrix threeChannelMatrix : threeChannelMatrixList) {
             List<Double> feature = new ArrayList<>();
-            MeanClustering meanClustering = new MeanClustering(sqNub);
+            MeanClustering meanClustering = new MeanClustering(sqNub, templeConfig);
             Matrix matrixR = threeChannelMatrix.getMatrixR();
             Matrix matrixG = threeChannelMatrix.getMatrixG();
             Matrix matrixB = threeChannelMatrix.getMatrixB();
@@ -165,7 +165,7 @@ public class Convolution extends Frequency {
         //System.out.println(matrixBD.getString());
     }
 
-    public List<Double> getCenterColor(ThreeChannelMatrix threeChannelMatrix, int poolSize, int sqNub) throws Exception {
+    public List<Double> getCenterColor(ThreeChannelMatrix threeChannelMatrix, int poolSize, int sqNub, TempleConfig templeConfig) throws Exception {
         Matrix matrixR = threeChannelMatrix.getMatrixR();
         Matrix matrixG = threeChannelMatrix.getMatrixG();
         Matrix matrixB = threeChannelMatrix.getMatrixB();
@@ -175,7 +175,7 @@ public class Convolution extends Frequency {
         RGBSort rgbSort = new RGBSort();
         int x = matrixR.getX();
         int y = matrixR.getY();
-        MeanClustering meanClustering = new MeanClustering(sqNub);
+        MeanClustering meanClustering = new MeanClustering(sqNub, templeConfig);
         for (int i = 0; i < x; i++) {
             for (int j = 0; j < y; j++) {
                 double[] color = new double[]{matrixR.getNumber(i, j), matrixG.getNumber(i, j), matrixB.getNumber(i, j)};

src/main/java/org/wlld/imageRecognition/MeanClustering.java

@@ -16,8 +16,9 @@ public class MeanClustering {
         return matrices;
     }
 
-    public MeanClustering(int speciesQuantity) {
+    public MeanClustering(int speciesQuantity, TempleConfig templeConfig) {
         this.speciesQuantity = speciesQuantity;//聚类的数量
+        size = templeConfig.getFood().getRegressionNub();
     }
 
     public void setColor(double[] color) throws Exception {

src/main/java/org/wlld/imageRecognition/Operation.java

@@ -111,7 +111,7 @@ public class Operation {//进行计算
             int times = templeConfig.getFood().getTimes();
             for (int i = 0; i < times; i++) {
                 List<Double> feature = convolution.getCenterColor(threeChannelMatrix1, templeConfig.getPoolSize(),
-                        templeConfig.getFeatureNub());
+                        templeConfig.getFeatureNub(), templeConfig);
                 if (templeConfig.isShowLog()) {
                     System.out.println(tag + ":" + feature);
                 }
@@ -178,7 +178,7 @@ public class Operation {//进行计算
             ThreeChannelMatrix threeChannelMatrix1 = convolution.getRegionMatrix(threeChannelMatrix, minX, minY, xSize, ySize);
             //convolution.filtering(threeChannelMatrix1);//光照过滤
             List<Double> feature = convolution.getCenterColor(threeChannelMatrix1, templeConfig.getPoolSize(),
-                    templeConfig.getFeatureNub());
+                    templeConfig.getFeatureNub(), templeConfig);
             if (templeConfig.isShowLog()) {
                 System.out.println(feature);
             }
@@ -272,7 +272,7 @@ public class Operation {//进行计算
                 CoverBody coverBody = new CoverBody();
                 Map<Integer, Double> tag = new HashMap<>();
                 tag.put(entry.getKey(), 1.0);
-                List<List<Double>> lists = convolution.kAvg(entry.getValue(), sqNub, regionSize);
+                List<List<Double>> lists = convolution.kAvg(entry.getValue(), sqNub, regionSize, templeConfig);
                 size = lists.size();
                 coverBody.setFeature(lists);
                 coverBody.setTag(tag);
@@ -296,7 +296,7 @@ public class Operation {//进行计算
         if (templeConfig.getStudyPattern() == StudyPattern.Cover_Pattern) {
             Map<Integer, Double> coverMap = new HashMap<>();
             Map<Integer, Integer> typeNub = new HashMap<>();
-            List<List<Double>> lists = convolution.kAvg(matrix, sqNub, regionSize);
+            List<List<Double>> lists = convolution.kAvg(matrix, sqNub, regionSize, templeConfig);
             //特征塞入容器完毕
             int size = lists.size();
             int all = 0;

src/main/java/org/wlld/imageRecognition/segmentation/Watershed.java

@@ -38,6 +38,7 @@ public class Watershed {
     private double columnMark;//列过滤
     private List<Specifications> specifications;//过滤候选区参数
     private List<RgbRegression> trayBody;//托盘参数
+    private double trayTh;
 
     public Watershed(ThreeChannelMatrix matrix, List<Specifications> specifications, TempleConfig templeConfig) throws Exception {
         if (matrix != null && specifications != null && specifications.size() > 0) {
@@ -61,6 +62,7 @@ public class Watershed {
             xSize = this.matrix.getX() / regionNub;
             ySize = this.matrix.getY() / regionNub;
             maxIou = templeConfig.getCutting().getMaxIou();
+            trayTh = templeConfig.getFood().getTrayTh();
             // System.out.println("xSize===" + xSize + ",ysize===" + ySize);
             rainfallMap = new Matrix(this.matrix.getX(), this.matrix.getY());
             regionMap = new Matrix(regionNub, regionNub);
@@ -77,7 +79,7 @@ public class Watershed {
                 matrixB.getNumber(x, y) / 255};
         for (RgbRegression rgbRegression : trayBody) {
             double dist = rgbRegression.getDisError(rgb);
-            if (dist < 0) {
+            if (dist < trayTh) {
                 isTray = true;
                 break;
             }
@@ -274,15 +276,15 @@ public class Watershed {
                 regionBodies.add(regionBody);
             }
         }
-        for (RegionBody regionBody : regionBodies) {
+//        for (RegionBody regionBody : regionBodies) {
-            int minX = regionBody.getMinX();
+//            int minX = regionBody.getMinX();
-            int maxX = regionBody.getMaxX();
+//            int maxX = regionBody.getMaxX();
-            int minY = regionBody.getMinY();
+//            int minY = regionBody.getMinY();
-            int maxY = regionBody.getMaxY();
+//            int maxY = regionBody.getMaxY();
-            System.out.println("minX==" + minX + ",minY==" + minY + ",maxX==" + maxX + ",maxY==" + maxY);
+//            System.out.println("minX==" + minX + ",minY==" + minY + ",maxX==" + maxX + ",maxY==" + maxY);
-        }
+//        }
-        //return iou(regionBodies);
+        return iou(regionBodies);
-        return regionBodies;
+        //return regionBodies;
     }
 
     private List<RegionBody> iou(List<RegionBody> regionBodies) {

src/main/java/org/wlld/param/Food.java

@@ -18,6 +18,15 @@ public class Food {
     private double columnMark = 0.25;//列痕迹过滤
     private List<RgbRegression> trayBody = new ArrayList<>();//托盘实体参数
     private int regressionNub = 10000;//回归次数
+    private double trayTh = 0.1;//托盘回归阈值
+
+    public double getTrayTh() {
+        return trayTh;
+    }
+
+    public void setTrayTh(double trayTh) {
+        this.trayTh = trayTh;
+    }
 
     public int getRegressionNub() {
         return regressionNub;