Shi-Tomasi角点检测
Harris算法原始的定义是将矩阵M的行列式值与M的迹相减,再将差值同预先给定的阈值进行比较
由于harris算法的稳定性和k值相关,因此在harris的基础上发现,角点的稳定性和矩阵M的较小特征值相关。
即可以直接计算出矩阵M的特征值,用较小的特征值与阈值比较,大于阈值的则为角点
Api说明:
代码:
#include<opencv2/opencv.hpp>
#include<math.h>
#include<iostream>
using namespace cv;
using namespace std;
const char* output_title = "HarrisCornerDetection";
int corn_num = 25;
int corn_max = 200;
Mat src, gray_src;
void Tomasi_Demo(int, void*);
int main(int argc, char** argv) {
src = imread("C:/Users/18929/Desktop/博客项目/项目图片/18.jpg");
if (src.empty()) {
printf("could not load image");
return -1;
}
imshow("input_img", src);
namedWindow(output_title, CV_WINDOW_AUTOSIZE);
//首先转换为灰度图像
cvtColor(src, gray_src, COLOR_BGR2GRAY);
createTrackbar("Threshold:", output_title, &corn_num, corn_max, Tomasi_Demo);
Tomasi_Demo(0, 0);
waitKey(0);
return 0;
}
void Tomasi_Demo(int, void*) {
//定义最小显示多少个corner
if (corn_num<5)
{
corn_num = 5;
}
//定义容器装corn检测结果
vector<Point2f> corners;
//定义参数
double qualityLevel = 0.01;
double minDistance = 10;
int blockSize = 3;
double useHarris = false;
double k = 0.04;
Mat resultImg = src.clone();
//交点检测
goodFeaturesToTrack(gray_src, corners, corn_num, qualityLevel, minDistance, Mat(), blockSize, useHarris, k);
printf("Number of Dected Corners:%d", corners.size());
//画出角点
for (size_t i = 0; i < corners.size(); i++)
{
circle(resultImg, corners[i], 2, Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255)), 2, 8, 0);
}
dst = Mat::zeros(gray_src.size(), CV_32FC1);
//定义参数
int blockSize = 2;
int ksize = 3;
double k = 0.04;
//角点检测
cornerHarris(gray_src, dst, blockSize, ksize, k, BORDER_DEFAULT);
//检测完后,通过归一化操作,将像素限制在0-255内
normalize(dst, norm_dst, 0, 255, NORM_MINMAX, CV_32FC1, Mat());
//增强画质,看的更清晰
convertScaleAbs(norm_dst, norm_result);
//画出检测结果
Mat resultImg = src.clone();
for (int row = 0; row < resultImg.rows; row++)
{
//获取对应结果图的行数
uchar* currentRow = norm_result.ptr(row);
for (int col = 0; col < resultImg.cols; col++)
{
int value = (int)*currentRow;
//比较,如果对应值笔阈值大,则为角点
if (value > thres) {
circle(resultImg, Point(col, row), 2, Scalar(0, 0, 255), 2, 8, 0);
}
currentRow++;
}
}
}