package ISQ.VSU.feature;

import java.io.BufferedReader;

import java.io.FileInputStream;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.util.ArrayList;
import java.util.Iterator;
import java.util.List;
import java.util.Map;

import ESU.array.Double_ESU;
import OEU.LYG4DQS4D.LYG9DWithDoubleTopSort4D;

//准备用肽展公式来设计个 皮肤病检测程序.
//罗瑶光 20210710
//软件思想 肽展腐蚀变换. 计算机视觉. 
//硬件工具, imageIO, javaCV. 像素头
//因为肽展公式和思想GPL2.0开源, 本程序源码同样GPL2.0开源.
//
//作者,著作权人： 罗瑶光，浏阳
public class SkinPathDetectTrip{
  public static List<Double[]> fileCells;
  public static List<String> fileNames;
  public static void main(String[] argv) throws IOException {
    //皮肤病图片识别
    //1 读取一张图片
    //String testImagePath= "C:\\Users\\yaogu\\Desktop\\project\\开源门户
    //\\deciphering\\F_DB.jpg";
    String testImagePath= "C:\\Users\\yaogu\\Desktop\\project\\开源门户"
        + "\\deciphering\\一般脏_白癜风伴白发-色素失禁症-毛发稀疏伴色素沉着.jpg";
    //2 计算图片训练值
    int 腐蚀浓度= 5;
    int[][][] getSkinBy= new GetSkinBy().getSkinBy(testImagePath);
    int[][][] doPDEofSkin= new DoPDEofSkin().doPDEofSkin_IDUQ(getSkinBy, 腐蚀浓度);
    //new ReadWritePng().writePNG("C:\\Users\\yaogu\\Desktop\\project\\开源门户
    //\\deciphering\\F_DB1.jpg", doPDEofSkin);
    int 像素阀值= 10; 
    int 像素比精度= 5; 
    int 像素差精度= 8;
    double[] getStatisticCount= new GetStatisticCount().getStatisticCount(doPDEofSkin
        , 像素阀值, 像素比精度, 像素差精度);
    //这个getStatisticCount地方可以优化为 像奥运会比赛那样,明显太多德杂色进行自动或者认为剔除
    //, 如字的黑色, 一些图片的红色.
    //剔除后也可以最小值剔除, 如 一些散落的灰色(高斯噪), 无特征的三原同位色阶 颜色, 图片颜色等.
    double[] getStatisticRatio= new GetStatisticRatio().getStatisticRatio(getStatisticCount);
    //
    //3 遍历疾病数据表.
    initDeciphring("C:\\Users\\yaogu\\Desktop\\project\\开源门户\\deciphering\\F_DB6.txt");
    //4 打分
    Double[] score= new Double[fileNames.size()];
    String[] nameScore= new String[fileNames.size()];
    int pcaScale= 25;
    int upcaScale= 15;
    int icaScale= 15;
    int ecaScale= 20;
    GetColorRatioScore.scoreDeciphring(score, nameScore, getStatisticRatio, fileCells
        , fileNames
        , pcaScale, upcaScale, icaScale, ecaScale);



    //
    腐蚀浓度= 95;
    getSkinBy= new GetSkinBy().getSkinBy(testImagePath);
    doPDEofSkin= new DoPDEofSkin().doPDEofSkin_IDUQ(getSkinBy, 腐蚀浓度);
    //new ReadWritePng().writePNG("C:\\Users\\yaogu\\Desktop\\project\\开源门户\\deciphering
    //\\F_DB1.jpg", doPDEofSkin);
    像素阀值= 10; 
    像素比精度= 8; 
    像素差精度= 8;
    getStatisticCount= new GetStatisticCount().getStatisticCount(doPDEofSkin, 像素阀值
        , 像素比精度, 像素差精度);
    //这个getStatisticCount地方可以优化为 像奥运会比赛那样,明显太多德杂色进行自动或者认为剔除
    //, 如字的黑色, 一些图片的红色.
    //剔除后也可以最小值剔除, 如 一些散落的灰色(高斯噪), 无特征的三原同位色阶 颜色, 图片颜色等.
    getStatisticRatio= new GetStatisticRatio().getStatisticRatio(getStatisticCount);
    //
    //3 遍历疾病数据表.
    initDeciphring("C:\\Users\\yaogu\\Desktop\\project\\开源门户\\deciphering\\F_DB5.txt");
    //4 打分
    Double[] score酸= new Double[fileNames.size()];
    GetColorRatioScore.scoreDeciphring(score酸, nameScore, getStatisticRatio, fileCells
        , fileNames, pcaScale, upcaScale, icaScale, ecaScale);


    //
    //
    腐蚀浓度= 5;
    getSkinBy= new GetSkinBy().getSkinBy(testImagePath);
    doPDEofSkin= new DoPDEofSkin().doPDEofSkin_AOPM_VECS_IDUQ_TXH_AC(getSkinBy, 腐蚀浓度);
    //new ReadWritePng().writePNG("C:\\Users\\yaogu\\Desktop\\project\\开源门户
    //\\deciphering\\F_DB_17.jpg", doPDEofSkin);
    像素阀值= 10; 
    像素比精度= 2; 
    像素差精度= 8;
    getStatisticCount= new GetStatisticCount().getStatisticCount(doPDEofSkin, 像素阀值
        , 像素比精度, 像素差精度);
    //这个getStatisticCount地方可以优化为 像奥运会比赛那样,明显太多德杂色进行自动或者认为剔除
    //, 如字的黑色, 一些图片的红色.
    //剔除后也可以最小值剔除, 如 一些散落的灰色(高斯噪), 无特征的三原同位色阶 颜色, 图片颜色等.
    getStatisticRatio= new GetStatisticRatio().getStatisticRatio(getStatisticCount);
    //
    //3 遍历疾病数据表.
    initDeciphring("C:\\Users\\yaogu\\Desktop\\project\\开源门户\\deciphering"
        + "\\F_DB6_jian_17.txt");
    //4 打分
    Double[] score_17= new Double[fileNames.size()];
    pcaScale= 1;
    upcaScale= 45;
    icaScale= 1;
    ecaScale= 1;  
    GetColorRatioScore.scoreDeciphring(score_17, nameScore, getStatisticRatio
        , fileCells, fileNames, pcaScale, upcaScale, icaScale, ecaScale);


    //
    腐蚀浓度= 95;
    getSkinBy= new GetSkinBy().getSkinBy(testImagePath);
    doPDEofSkin= new DoPDEofSkin().doPDEofSkin_AOPM_VECS_IDUQ_TXH_AC(getSkinBy, 腐蚀浓度);
    //new ReadWritePng().writePNG("C:\\Users\\yaogu\\Desktop\\project\\开源门户
    //\\deciphering\\F_DB_17.jpg", doPDEofSkin);
    像素阀值= 10; 
    像素比精度= 3;//采样是5 
    像素差精度= 8;
    getStatisticCount= new GetStatisticCount().getStatisticCount(doPDEofSkin
        , 像素阀值, 像素比精度, 像素差精度);
    //这个getStatisticCount地方可以优化为 像奥运会比赛那样,明显太多德杂色进行自动或者认为剔除
    //, 如字的黑色, 一些图片的红色.
    //剔除后也可以最小值剔除, 如 一些散落的灰色(高斯噪), 无特征的三原同位色阶 颜色, 图片颜色等.
    getStatisticRatio= new GetStatisticRatio().getStatisticRatio(getStatisticCount);
    //
    //3 遍历疾病数据表.
    initDeciphring("C:\\Users\\yaogu\\Desktop\\project\\开源门户\\deciphering"
        + "\\F_DB6_suan_17.txt");
    //4 打分
    Double[] score_酸17= new Double[fileNames.size()];
    pcaScale= 1;
    upcaScale= 45;
    icaScale= 1;
    ecaScale= 1;  
    GetColorRatioScore.scoreDeciphring(score_酸17, nameScore, getStatisticRatio
        , fileCells, fileNames, pcaScale, upcaScale, icaScale, ecaScale);
    //融合
    for(int i= 0; i< score酸.length; i++) {
      //score[i]+= score酸[i]+ score_17[i]+ score_酸17[i]; 
      score[i]= score[i]* 1+ score酸[i]* 0.001+ score_17[i]* 1+ score_酸17[i]* 1; 
    }
    //5 筛选
    double[] scoreDouble= new double[score.length];
    for(int i= 0; i< scoreDouble.length; i++) {
      scoreDouble[i]= score[i];//
    }
    //改成map
    Map<Double, Map<String, Boolean>> map= Double_ESU.getMapFromDoubleStringArray(score
        , nameScore);  
    int 递归深度= 70; //避免同值冗余内存高峰
    int 堆栈广度= 7; //避免堆栈浪费计算高峰
    new LYG9DWithDoubleTopSort4D().sort(scoreDouble, 堆栈广度, 递归深度);
    //new Quick9DLYGWithString_ESU().sort(scoreInt, nameScore);

    //6 推荐
    Here:
      for(int i= 0; i< scoreDouble.length; i++) {
        if(!map.containsKey(scoreDouble[i])) {
          i++;
          continue Here;
        }
        Iterator<String> iterator= map.get(scoreDouble[i]).keySet().iterator();
        while(iterator.hasNext()) {
          String string= iterator.next();
          if(string.contains("白癜风伴")) {
            System.out.println("相似图片:"+ i+ "位"+ string+ "-----分数:"+ scoreDouble[i]);
          }else{
            if(i< 20) {
              System.out.println("相似图片:"+ i+ "位"+ string+ "-----分数:"+ scoreDouble[i]);
            }
          }
        }
        map.remove(scoreDouble[i]);
        //  System.out.println("相似图片:"+ nameScore[i]+ "-----分数:"+ scoreInt[i]);
      }
  }

  private static void initDeciphring(String string) throws NumberFormatException, IOException {
    fileNames= new ArrayList<>();
    fileCells= new ArrayList<>();
    String ctempString;
    InputStream inputStream= new FileInputStream(string);
    BufferedReader cReader= new BufferedReader(new InputStreamReader(inputStream, "GBK"));
    while ((ctempString= cReader.readLine())!= null) {  
      String[] lineCell= ctempString.split(">d>");
      if(lineCell.length> 1) {
        fileNames.add(lineCell[0]);
        Double[] lineDouble= new Double[lineCell.length- 1]; 
        for(int i= 0; i< lineDouble.length; i++) {
          lineDouble[i]= Double.valueOf(lineCell[i+ 1]);
        }
        fileCells.add(lineDouble);
      }
    }
    cReader.close();
  }
}