opencv实现图像细化效果

我不是女神ヾ 2022-08-10 06:18 235阅读 0赞

在图像处理中，有时候我们会想要提取图像的骨架，这是就需要对图像进行细化，opencv中没有直接进行细化的算法，网上大部分的细化算法都是基于以前IplImage结构的，对于想要使用新的C++接口的Mat结构需要进行一定的修改，本文的细化方法是基于Mat数据结构的，使用的是OpenCV2.4.9版本。  
参考自：[http://blog.csdn.net/qianchenglenger/article/details/19332011][http_blog.csdn.net_qianchenglenger_article_details_19332011]

//提取图像的骨架
    void ImgThin(cv::Mat src,int maxIterations=-1)
    {
        if (src.empty()) return;//图像为空，直接返回
        cv::threshold(src, src, m_dThreshold, 1, CV_THRESH_BINARY);//转为0或1的图像
    
        int ImgHeight = src.rows;
        int ImgWidth = src.cols;
    
        int count = 0;  //记录迭代次数
        while (true)
        {
            count++;
            if (maxIterations != -1 && count > maxIterations) //限制次数并且迭代次数到达
                break;
    
            vector<pair<int, int> > mFlag; //用于标记需要删除的点
            //对点标记
            for (int i = 0; i < ImgHeight; ++i)
            {
                for (int j = 0; j < ImgWidth; ++j)
                {
                    //如果满足四个条件，进行标记
                    // p9 p2 p3
                    // p8 p1 p4
                    // p7 p6 p5
                    int p1 = src.at<uchar>(i, j);
                    int p2 = (i == 0) ? 0 : src.at<uchar>(i - 1, j);
                    int p3 = (i == 0 || j == ImgWidth - 1) ? 0 : src.at<uchar>(i - 1, j + 1);
                    int p4 = (j == ImgWidth - 1) ? 0 : src.at<uchar>(i, j + 1);
                    int p5 = (i == ImgHeight - 1 || j == ImgWidth - 1) ? 0 : src.at<uchar>(i + 1, j + 1);
                    int p6 = (i == ImgHeight - 1) ? 0 : src.at<uchar>(i + 1, j);
                    int p7 = (i == ImgHeight - 1 || j == 0) ? 0 : src.at<uchar>(i + 1, j - 1);
                    int p8 = (j == 0) ? 0 : src.at<uchar>(i, j - 1);
                    int p9 = (i == 0 || j == 0) ? 0 : src.at<uchar>(i - 1, j - 1);
    
                    if ((p2 + p3 + p4 + p5 + p6 + p7 + p8 + p9) >= 2 && (p2 + p3 + p4 + p5 + p6 + p7 + p8 + p9) <= 6)
                    {
                        int ap = 0;
                        if (p2 == 0 && p3 == 1) ++ap;
                        if (p3 == 0 && p4 == 1) ++ap;
                        if (p4 == 0 && p5 == 1) ++ap;
                        if (p5 == 0 && p6 == 1) ++ap;
                        if (p6 == 0 && p7 == 1) ++ap;
                        if (p7 == 0 && p8 == 1) ++ap;
                        if (p8 == 0 && p9 == 1) ++ap;
                        if (p9 == 0 && p2 == 1) ++ap;
    
                        if (ap == 1)
                        {
                            if (p2*p4*p6 == 0)
                            {
                                if (p4*p6*p8 == 0)
                                {
                                    //标记
                                    mFlag.push_back(make_pair(i, j));
                                }
                            }
                        }
                    }
                }
            }
    
            //将标记的点删除
            for (vector<pair<int, int> >::iterator i = mFlag.begin(); i != mFlag.end(); ++i)
            {
                src.at<uchar>(i->first, i->second) = 0;
            }
    
            //直到没有点满足，算法结束
            if (mFlag.size() == 0) break;
            else mFlag.clear();//将mFlag清空
    
            //对点标记
            for (int i = 0; i < ImgHeight; ++i)
            {
                for (int j = 0; j < ImgWidth; ++j)
                {
                    //如果满足四个条件，进行标记
                    // p9 p2 p3
                    // p8 p1 p4
                    // p7 p6 p5
                    int p1 = src.at<uchar>(i, j);
                    if (p1 != 1) continue;
                    int p2 = (i == 0) ? 0 : src.at<uchar>(i - 1, j);
                    int p3 = (i == 0 || j == ImgWidth - 1) ? 0 : src.at<uchar>(i - 1, j + 1);
                    int p4 = (j == ImgWidth - 1) ? 0 : src.at<uchar>(i, j + 1);
                    int p5 = (i == ImgHeight - 1 || j == ImgWidth - 1) ? 0 : src.at<uchar>(i + 1, j + 1);
                    int p6 = (i == ImgHeight - 1) ? 0 : src.at<uchar>(i + 1, j);
                    int p7 = (i == ImgHeight - 1 || j == 0) ? 0 : src.at<uchar>(i + 1, j - 1);
                    int p8 = (j == 0) ? 0 : src.at<uchar>(i, j - 1);
                    int p9 = (i == 0 || j == 0) ? 0 : src.at<uchar>(i - 1, j - 1);
    
                    if ((p2 + p3 + p4 + p5 + p6 + p7 + p8 + p9) >= 2 && (p2 + p3 + p4 + p5 + p6 + p7 + p8 + p9) <= 6)
                    {
                        int ap = 0;
                        if (p2 == 0 && p3 == 1) ++ap;
                        if (p3 == 0 && p4 == 1) ++ap;
                        if (p4 == 0 && p5 == 1) ++ap;
                        if (p5 == 0 && p6 == 1) ++ap;
                        if (p6 == 0 && p7 == 1) ++ap;
                        if (p7 == 0 && p8 == 1) ++ap;
                        if (p8 == 0 && p9 == 1) ++ap;
                        if (p9 == 0 && p2 == 1) ++ap;
    
                        if (ap == 1)
                        {
                            if (p2*p4*p8 == 0)
                            {
                                if (p2*p6*p8 == 0)
                                {
                                    //标记
                                    mFlag.push_back(make_pair(i, j));
                                }
                            }
                        }
                    }
                }
            }
            //删除
            for (vector<pair<int, int> >::iterator i = mFlag.begin(); i != mFlag.end(); ++i)
            {
                src.at<uchar>(i->first, i->second) = 0;
            }
    
            //直到没有点满足，算法结束
            if (mFlag.size() == 0) break;
            else mFlag.clear();//将mFlag清空
        }
    
        cv::threshold(src, src, 0, 255, CV_THRESH_BINARY);//二值化图像
    }

对图像进行测试：

int main(int argc, char*argv[])  
    { 
        cv::Mat src = imread("test.jpg");
        imshow("src",src);
        ImgThin(src);
        imshow("thin",src);
        system("pause");
        return 0;
    }

![这里写图片描述][20151114162127000]

![这里写图片描述][20151114162138656]

[http_blog.csdn.net_qianchenglenger_article_details_19332011]: http://blog.csdn.net/qianchenglenger/article/details/19332011
[20151114162127000]: /images/20220731/930efdc29fd84e939cb6b9ab46439763.png
[20151114162138656]: /images/20220731/aa2477b2b4a84361b2d51adad9001ac4.png