使用opencv中基于高斯混合模型（GMM）的EM算法进行图像分割

环境

Win10+VS2015+opencv3.4.x

demo源码：

//图形分割
void segment(Mat img) {
	namedWindow("srcImg", 0);
	imshow("srcImg", img);
	int wid = img.cols;
	int hig = img.rows;
	int dim = img.channels();
	int sampleCount = wid*hig;
	int clusterCount = 3;
	//颜色索引表
	Scalar colorTab[5] = {
		Scalar(0,255,255),
		Scalar(0,255,0),
		Scalar(255,0,0),
		Scalar(255,0,255),
		Scalar(255,255,0),

	};
	Mat points(sampleCount, dim, CV_32FC1, Scalar(10)); //sampleCount行，dim列
 

	int index = 0;
	//图像转数据点
	for (int i=0;i<hig;i++)
	{
		for (int j=0;j<wid;j++)
		{
			index = i*wid + j;
			Vec3b bgr = img.at<Vec3b>(i, j);

			points.at<float>(index, 0) = static_cast<int>(bgr[0]);
			points.at<float>(index, 1) = static_cast<int>(bgr[1]);
			points.at<float>(index, 2) = static_cast<int>(bgr[2]);
		}
	}

	//EM
	Mat lables;
	Ptr<EM> em_model = EM::create();
	em_model->setClustersNumber(clusterCount);
	em_model->setCovarianceMatrixType(EM::COV_MAT_SPHERICAL);  // 协方差矩阵
	em_model->setTermCriteria(TermCriteria(TermCriteria::EPS + TermCriteria::COUNT, 100, 0.1));//停止条件
	em_model->trainEM(points, noArray(), lables, noArray());
	//结果点集转图像
	Mat segImg = Mat::zeros(img.size(), img.type());
	Mat sample(dim, 1, CV_32FC1);
 
	for (int i=0;i<hig;i++)
	{
		for (int j=0;j<wid;j++)
		{
			 index= i*wid + j;
			int lable = lables.at<int>(index, 0); 
			Scalar color = colorTab[lable];
 
			Vec3b vec(colorTab[lable][0], colorTab[lable][1], colorTab[lable][2]);
			segImg.at<Vec3b>(i, j) = vec;
		}
	}

	namedWindow("segmentation", 0);
	imshow("segmentation", segImg);
	waitKey(0);
}

速度有点儿慢。。。耐心等

结果

图片来源：https://timgsa.baidu.com/timg?image&quality=80&size=b9999_10000&sec=1558871961266&di=287e6143f8bfdca5933a7509d126a24a&imgtype=0&src=http%3A%2F%2Ftva1.sinaimg.cn%2Fcrop.0.2.1242.1242.1024%2F73fe846fjw8f9yk2xee35j20yi0ym410.jpg
 

对比另一篇博客，使用opencv中K-Means方法进行基于像素值的图像分割和背景替换

https://blog.csdn.net/ganwenbo2011/article/details/90577122