本文主要是介绍基于opencv的超声扇形区域检测(C++),希望对大家解决编程问题提供一定的参考价值,需要的开发者们随着小编来一起学习吧!
简介
从超声采集的图像会带有部分用户操作界面的组件,若使用全局图片进行病灶检测,操作界面的组件可能会对检测结果造成干扰,通过扇形区域检测可防止这种情况的出现。
Code
LargestConnecttedComponent函数出自该博客《opencv 获取图像最大连通域 c++和python版》
// 获取检测区域边界
cv::Rect GetBorder(cv::Mat img) {cv::Mat gray;// rgb转灰度图cvtColor(img, gray, cv::COLOR_BGR2GRAY);int counts[256] = { 0 };int value;// 统计0~255各颜色出现次数for (int row_id = 0; row_id < gray.rows; row_id++) {for (int col_id = 0; col_id < gray.cols; col_id++){counts[(int)gray.at<uchar>(row_id, col_id)] += 1;}}// 找出出现次数最多的颜色int max_val = 0;int max_index = 0;for (int i = 0; i < 256; i++){if (counts[i] > max_val) {max_index = i;max_val = counts[i];}}cv::Mat thresh, morph_open_mat, morph_close_mat;// 阈值处理threshold(gray, thresh, max_index, 255, cv::THRESH_BINARY);// 定义kernelcv::Mat kernel = cv::getStructuringElement(cv::MORPH_RECT, cv::Size(1, 5));// 开运算morphologyEx(thresh, morph_open_mat, cv::MORPH_OPEN, kernel);// 闭运算morphologyEx(morph_open_mat, morph_close_mat, cv::MORPH_CLOSE, kernel);cv::Mat lcc_mat;// 最大连通区域检测LargestConnecttedComponent(morph_close_mat, lcc_mat);int max_x = 0, max_y = 0, min_x = gray.cols, min_y = gray.rows;// 找到矩形角点for (int row_id = 0; row_id < gray.rows; row_id++) {for (int col_id = 0; col_id < gray.cols; col_id++){if ((int)lcc_mat.at<uchar>(row_id, col_id) == 255) {if (col_id > max_x) {max_x = col_id;}if (col_id < min_x) {min_x = col_id;}if (row_id > max_y) {max_y = row_id;}if (row_id < min_y) {min_y = row_id;}}}}// 返回角点cv::Rect rect(cv::Point(min_x, min_y), cv::Point(max_x, max_y));return rect;
}
效果
这篇关于基于opencv的超声扇形区域检测(C++)的文章就介绍到这儿,希望我们推荐的文章对编程师们有所帮助!