基于OpenCV-python的夜间灯光识别源码+使用文档+全部资料（优秀项目）.zip

#include "lamps_detection.h" #include <vector> using namespace std; CLampsDetection::CLampsDetection() { m_nStructElementSize = 2; } void CLampsDetection::LampsDetectProcess(cv::Mat& img,long long time_stamp,std::vector<Object>& detections_list) { if(img.empty()) { return; } /*cv::cvtColor(img,m_image_gray,CV_BGR2GRAY); //cv::resize(m_image_gray,m_image_gray,cv::Size(m_imgW,m_imgH)); cv::threshold(m_image_gray,m_image_gray , 150, 255, cv::THRESH_BINARY); cv::Mat element = getStructuringElement(cv::MORPH_ELLIPSE, cv::Size(2 * m_nStructElementSize + 1, 2 * m_nStructElementSize + 1), cv::Point(m_nStructElementSize, m_nStructElementSize)); cv::erode(m_image_gray, m_image_gray, element); cv::dilate(m_image_gray, m_image_gray, element); */ cv::SimpleBlobDetector::Params pDefaultBLOB; //二值化 pDefaultBLOB.minThreshold = 150; //二值化的起始阈值 40 pDefaultBLOB.maxThreshold = 255; pDefaultBLOB.thresholdStep = 2; pDefaultBLOB.minRepeatability = 1; pDefaultBLOB.minDistBetweenBlobs = 10; //斑点颜色 pDefaultBLOB.filterByColor = true; pDefaultBLOB.blobColor = 255; //面积控制 pDefaultBLOB.filterByArea = true; pDefaultBLOB.minArea = 1000; pDefaultBLOB.maxArea = 3000000; //形状圆 pDefaultBLOB.filterByCircularity = true; pDefaultBLOB.minCircularity = 0.3f; pDefaultBLOB.maxCircularity = std::numeric_limits<float>::max(); //凹 pDefaultBLOB.filterByConvexity = true; pDefaultBLOB.minConvexity = 0.8; //0.05f//斑点的最小凹度 pDefaultBLOB.maxConvexity = std::numeric_limits<float>::max(); //斑点的最大凸度 //惯性比 pDefaultBLOB.filterByInertia = true; pDefaultBLOB.minInertiaRatio = 0.3f; //斑点的最小惯性率 圆的值为1，直线的值为0 //.05f pDefaultBLOB.maxInertiaRatio = 1; //斑点的最大惯性率 //凸 pDefaultBLOB.filterByConvexity = true; pDefaultBLOB.minConvexity = 0.3f; pDefaultBLOB.maxConvexity = std::numeric_limits<float>::max(); //*用参数创建对象 cv::Ptr<cv::SimpleBlobDetector> blob=cv::SimpleBlobDetector::create(pDefaultBLOB); vector<cv::KeyPoint> key_points; blob->detect(img,key_points); cv::Mat outImg; static int i=0; if(key_points.size()>0) { for(int i=0;i<key_points.size();i++) { cout << "Blob_size 直径：" << key_points[i].size << endl; cout << "Blob_response 点强度:" << key_points[i].response << endl; cout << "Blob_pt 点位置：" << key_points[i].pt << endl; cout << "keypoints _angle" << key_points[i].angle << endl; cout << "keypoints_octave" << key_points[i].octave << endl; } cout<<"size ="<<key_points.size()<<endl; outImg =img.clone(); //绘制结果 cv::drawKeypoints(img,key_points,outImg,cv::Scalar(0,0,255),cv::DrawMatchesFlags::DRAW_RICH_KEYPOINTS); char str[50]; sprintf(str,"%d.jpg",i); i++; imwrite(str,outImg); cout<<str<<endl; //sprintf(str,"---%d.jpg",i); //imwrite(str,m_image_gray); } }