opencv 模闆比對 matchTemplate

opencv 模闆比對 matchTemplate

demo：http://download.csdn.net/detail/keen_zuxwang/9847695

模闆比對

在一幅圖像中尋找一個特定目标的方法。周遊圖像中的每一個位置，比較與模闆是否“相似”，當相似度足夠高時，就認為找到了目标。

常用于目标檢測、相似度分析

//! computes the proximity map for the raster template and the image where the template is searched for
CV_EXPORTS_W void matchTemplate( InputArray image, InputArray templ,
                                 OutputArray result, int method );
           

功能：在輸入搜尋圖像中按照method方法比對模闆

image

用于搜尋的輸入圖像, 8U 或 32F, 大小 W-H

templ

比對模闆圖像，和image類型相同， 大小 w-h

result

比對結果圖像, 類型 32F, 大小 (W-w+1)-(H-h+1)

method

比對方法，根據實際情況而定：

TM_SQDIFF：

平方差進行比對，最佳的比對結果在結果為0處，值越大比對結果越差

TM_SQDIFF_NORMED：

歸一化的平方差進行比對，最佳比對在結果為0處

TM_CCORR：

相關性比對方法，該方法使用源圖像與模闆圖像的卷積結果進行比對，最佳比對位置在值最大處，值越小比對結果越差

TM_CCORR_NORMED：

歸一化的相關性比對方法，與相關性比對方法類似，最佳比對位置也是在值最大處

TM_CCOEFF：

相關性系數比對方法，該方法使用源圖像與其均值的差、模闆與其均值的差二者之間的相關性進行比對，最佳比對結果在值等于1處，最差比對結果在值等于-1處，值等于0直接表示二者不相關

TM_CCOEFF_NORMED：

歸一化的相關性系數比對方法，正值表示比對的結果較好，負值則表示比對的效果較差，也是值越大，比對效果也好

平方差比對到相關系數比對，比對的準确度越來越高(計算代價也随之升高). 通常可測試實驗,選擇同時兼顧速度和精度的最佳比對方法

//! finds global minimum and maximum array elements and returns their values and their locations
CV_EXPORTS_W void minMaxLoc(InputArray src, CV_OUT double* minVal,
                           CV_OUT double* maxVal=, CV_OUT Point* minLoc=,
                           CV_OUT Point* maxLoc=, InputArray mask=noArray());
           

JNI端程式：

tempray.cpp：

#include <string>  
#include <jni.h>
#include <android/log.h>
#include <iostream>
#include <fstream>
#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/imgproc/imgproc.hpp>
#include <opencv2/objdetect/objdetect.hpp>
#include <opencv2/ml/ml.hpp>

#ifdef __cplusplus
extern "C" {
#endif

//#define LOG_TAG "show infomation"
#define Template 100

#define LOG_TAG "libtracking"
#define  LOGI(...)  __android_log_print(ANDROID_LOG_INFO,LOG_TAG,__VA_ARGS__)
#define  LOGD(...)  __android_log_print(ANDROID_LOG_DEBUG,LOG_TAG,__VA_ARGS__)
#define  LOGE(...)  __android_log_print(ANDROID_LOG_ERROR,LOG_TAG,__VA_ARGS__)

using namespace cv;  
using namespace std;  

int result_cols;
int result_rows;
double minVal;
double maxVal;
Point minLoc;
Point maxLoc;
Point matchLoc;
int match_method = TM_CCORR;//CV_TM_SQDIFF;

JNIEXPORT void JNICALL Java_org_opencv_samples_tutorial2_Tutorial2Activity_grayProc(JNIEnv* env, jclass obj, jlong imageGray,
        jint touch_x,jint touch_y)
{

    int width,height;
    int k = ,n = ,i,j;
    CvScalar s;

    Mat result;
    Mat mat = Mat(*((Mat*)imageGray));
    width = mat.rows;
    height = mat.cols;

    result_cols = height - Template + ;
    result_rows = width - Template + ;

    cv::Mat img = cv::Mat(Template,Template,CV_8UC3,cv::Scalar(,,));
    IplImage pI = mat;
    IplImage pI_2 = img;

    Mat img_display = cv::Mat(width,height,CV_8UC3,cv::Scalar(,,));
    IplImage pI_3 = img_display;
    for(i=;i<width;i++){
        for(j=;j<height;j++){
            s = cvGet2D(&pI,i,j);
            cvSet2D(&pI_3,i,j,s);
        }
    }

    if((touch_x < (Template/)) || (touch_x>(height-(Template/))) || (touch_y < (Template/)) || (touch_y > (width-(Template/)))){
        //__android_log_print(ANDROID_LOG_ERROR, "JNITag","touch_x,touch_y is too small or too large\n");
        LOGD("touch_x = %d, touch_y = %d, width = %d, height = %d", touch_x, touch_y, width, height);
        return;
    }

    for(i=(touch_x-(Template/));i<(touch_x+(Template/));i++){
        for(j=(touch_y-(Template/));j<(touch_y+(Template/));j++){
            s = cvGet2D(&pI,j,i);
            cvSet2D(&pI_2,k,n,s);
            n++;
        }   
        k++;
        n=;
    }

    /*
    result.create(result_cols, result_rows, CV_32FC1 );// 模闆比對存儲Mat
    */

    /// Do the Matching and Normalize
    matchTemplate( img_display, img, img_display, match_method);
    normalize(img_display, img_display, , , NORM_MINMAX, -, Mat() );
    /// Localizing the best match with minMaxLoc

    minMaxLoc(img_display, &minVal, &maxVal, &minLoc, &maxLoc, Mat() );
    /// For SQDIFF and SQDIFF_NORMED, the best matches are lower values. For all the other methods, the higher the better
    if(match_method == CV_TM_SQDIFF || match_method == CV_TM_SQDIFF_NORMED ){
        matchLoc = minLoc;
    }else{
        matchLoc = maxLoc;
    }

    rectangle(mat, matchLoc, Point(matchLoc.x + img.cols , matchLoc.y + img.rows ), Scalar::all(), , , );

}

int run_time;

JNIEXPORT jlong JNICALL Java_org_opencv_samples_tutorial2_Tutorial2Activity_grayProc0(JNIEnv* env, jclass obj, jlong imageGray, jlong tempGray)
{
    Mat img_display =  Mat(*(Mat*)imageGray); //Mat: JAVA->JNI, 取位址
    Mat img  =  Mat(*(Mat*)tempGray);

    Mat input;
    input.create(img_display.size(), CV_8UC3);
    cvtColor(img_display, input, CV_BGRA2BGR);

    /// Do the Matching and Normalize，按NORM_MINMAX進行比對後的歸一化
    matchTemplate( img_display, img, img_display, match_method);
    normalize(img_display, img_display, , , NORM_MINMAX, -, Mat()); 

    // Localizing the best match with minMaxLoc，模闆在原圖像中比對的最大/最小值點及位置
    minMaxLoc(img_display, &minVal, &maxVal, &minLoc, &maxLoc, Mat());

    // For SQDIFF and SQDIFF_NORMED, the best matches are lower values. For all the other methods, the higher the better
    if(match_method == CV_TM_SQDIFF || match_method == CV_TM_SQDIFF_NORMED ){
        matchLoc = minLoc;
    }else{
        matchLoc = maxLoc;
    }

    LOGD("matchLoc.x = %d, matchLoc.y = %d, img.cols = %d, img.rows = %d", matchLoc.x, matchLoc.y, img.cols, img.rows);

    //框出查找到的原圖像中比對的模闆圖像
    rectangle(input, matchLoc, Point(matchLoc.x + img.cols , matchLoc.y + img.rows ), Scalar(,,), , , );//Scalar::all(0)

    Mat *mat = new Mat(input);
    return (jlong) mat; // 傳回位址，Mat: JNI->JAVA
}

#ifdef __cplusplus
}
#endif
           

TM_CCOEFF：

TM_SQDIFF：