基于opencv的缺陷检测

1..基于opencv 差分图像二值分析的刀片缺陷检测

#include 
#include 
using namespace cv;
using namespace std;
string rootdir = "F:/Study/opencv/opencv-4.8.0/opencv/sources/samples/data/";
void sort_box(vector &boxes) {
	int size = boxes.size();
	for (int i = 0; i  rects, vector &defect, Mat &tpl) {
	int h = tpl.rows;
	int w = tpl.cols;
	int size = rects.size();
	for (int i = 0; i (row, col);
				if (pv == 255) {
					count++;
				}
			}
		}

		// 填充一个像素宽
		int mh = mask.rows + 2;
		int mw = mask.cols + 2;
		Mat m1 = Mat::zeros(Size服务器托管网(mw, mh), mask.type());
		Rect mroi;
		mroi.x = 1;
		mroi.y = 1;
		mroi.height = mask.rows;
		mroi.width = mask.cols;
		mask.copyTo(m1(mroi));服务器托管网
		/*imshow("mask2", mask);
		imshow("ml", m1);*/
		waitKey(0);
		// 轮廓分析
		vector> contours;
		vector hierarchy;
		findContours(m1, contours, hierarchy, RETR_LIST, CHAIN_APPROX_SIMPLE);
		bool find = false;
		for (size_t t = 0; t  4.0 && (rect.y  10) {
				printf("ratio : %.2f, area : %.2f n", ratio, area);
				find = true;
			}
		}

		if (count > 50 && find) {
			printf("count : %d n", count);
			defect.push_back(rects[i]);
		}
	}
}

void multiple_defects_detection(Mat &src) {
	Mat tpl = imread(rootdir + "dt.png", IMREAD_GRAYSCALE);

	// 图像二值化
	Mat gray, binary;
	cvtColor(src, gray, COLOR_BGR2GRAY);
	threshold(gray, binary, 0, 255, THRESH_BINARY_INV | THRESH_OTSU);
	imshow("binary", binary);
	imwrite("D:/binary.png", binary);

	// 定义结构元素
	Mat se = getStructuringElement(MORPH_RECT, Size(3, 3), Point(-1, -1));
	morphologyEx(binary, binary, MORPH_OPEN, se);

	// 轮廓发现
	vector> contours;
	vector hierarchy;
	vector rects;
	findContours(binary, contours, hierarchy, RETR_LIST, CHAIN_APPROX_SIMPLE);
	int height = src.rows;
	for (size_t t = 0; t  (height / 2)) {
			continue;
		}
		if (area  defects;
	detect_defect(binary, rects, defects, tpl);

	// 显示检测结果
	for (int i = 0; i 

2.基于opencv DNN模块的UNet道路裂纹检测

#include
#include
using namespace cv;
using namespace std;
string model_dir="F:/Study/opencv/opencv-4.8.0/opencv/sources/samples/data/";
void resnet_surface_detection(Mat &image) {
	String defect_labels[] = { "In","Sc","Cr","PS","RS","Pa" };
	dnn::Net net = dnn::readNetFromONNX(model_dir + "surface_defect_resnet18.onnx");
	Mat inputBlob = dnn::blobFromImage(image, 0.00392, Size(200, 200), Scalar(127, 127, 127), false, false);
	inputBlob /= 0.5;

	// 执行图像分类
	Mat prob;
	net.setInput(inputBlob);
	prob = net.forward();

	// 得到最可能分类输出
	Mat probMat = prob.reshape(1, 1);
	Point classNumber;
	double classProb;
	minMaxLoc(probMat, NULL, &classProb, NULL, &classNumber);
	int classidx = classNumber.x;
	printf("n current image classification : %s, possible : %.2fn", defect_labels[classidx].c_str(), classProb);

	// 显示文本
	putText(image, defect_labels[classidx].c_str(), Point(20, 40), FONT_HERSHEY_SIMPLEX, 1.0, Scalar(0, 0, 255), 2, 8);
	imshow("基于分类的缺陷检测", image);
	waitKey(0);
}
int main() {
	Mat image = imread("F:/Study/opencv/opencv-4.8.0/opencv/sources/samples/data/NEU-CLS/NEU-CLS/Ps_1.bmp");
	resnet_surface_detection(image);
	return 0;
}

服务器托管，北京服务器托管，服务器租用 http://www.fwqtg.net

相关推荐: SQL手工注入

两要素 用户能够控制输入的内容 web应用把用户输入的内容，在没有经过过滤或者严格过滤的情况下带入到数据库中执行 分类 GET和POST 整数型，字符型，搜索型 万能密码 ’1 or 1 = 1# 1 or 1 = 1# 注释符： — (后面有空格） –+…