java opencv 图片放大缩小，角点检测和边缘检测（14）

谁践踏了优雅 2022-12-06 01:15 196阅读 0赞

# 什么是高斯金字塔 #

所谓的高斯金字塔是利用算法将图片放大，如下面的图形：

![20131227143130718][]

# openCv对应的函数： #

![20200911153056234.png][]

src输入图像

dst 目标图像

dstsize 放大尺寸

borderType 边框类型

# 实例: #

默认的尺度和边框

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70][]

src.size().toString():280x186  
dst.size().toString():560x372

# pyrDown #

和pyrUp对应的是pyrDown将图缩小

String filename = "D:\\360MoveData\\Users\\lxn\\Desktop\\opencvtest\\1.jpg";
    		Mat src = Imgcodecs.imread(filename);// 加载需要被蒙太奇的图片，原图
    		Mat dst = new Mat();
    		Imgproc.pyrDown(src, dst);
    		System.out.println("src.size().toString():"+src.size().toString());
    		System.out.println("dst.size().toString():"+dst.size().toString());
    		HighGui.imshow("原图", src);
    	    HighGui.imshow("高斯金字塔",dst);
    		HighGui.waitKey(0);

src.size().toString():280x186  
dst.size().toString():140x93

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 1][]

# 均值漂移 #

![20200911161237384.png][]

**src** The source 8-bit, 3-channel image.

**dst** The destination image of the same format and the same size as the source.

**sp** The spatial window radius. 空间窗口半径

**sr** The color window radius. 颜色窗口半径

String filename = "D:\\360MoveData\\Users\\lxn\\Desktop\\opencvtest\\1.jpg";
    		Mat src = Imgcodecs.imread(filename);// 加载需要被蒙太奇的图片，原图
    		Mat dst = new Mat();
    		
    		Imgproc.pyrUp(src, src);
    		
    		
    		Imgproc.pyrMeanShiftFiltering(src, dst, 10, 20);
    
    		HighGui.imshow("原图", src);
    	    HighGui.imshow("漂移均值图像",dst);
    		HighGui.waitKey(0);

具体的效果：

当ps=10 pr=20的时候，已经能看出区别

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 2][]

当ps=10 pr=60的时候

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 3][]

当ps=60 pr=10时：和原图变化不大

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 4][]

主要是颜色的半径，对整个图影响比较大

# 角点检测 #

![20200911164140886.png][]

String filename = "D:\\360MoveData\\Users\\lxn\\Desktop\\opencvtest\\1.jpg";
    		Mat src = Imgcodecs.imread(filename);// 加载需要被蒙太奇的图片，原图
    		Mat dst = new Mat();
    		
    		Imgproc.pyrUp(src, src);
    		
    		Imgproc.cvtColor(src, dst, Imgproc.COLOR_BGR2GRAY);
    		
    		Mat img=new Mat();
    		
    		Imgproc.preCornerDetect(dst, img,3);    
    	    float[] floats = new float[img.cols()];
    	    for (int i = 0; i < img.rows() ; i++) {
    	    	img.get(i,0,floats);
    	        for (int j = 0; j < floats.length; j++) {
    	            if (floats[j] > 0.0005){//越接近于角点数值越大
    	                Imgproc.circle(src,new Point(j,i),1,new Scalar(255,0,255));
    	            }
    	        }
    	    }
    		HighGui.imshow("原图", src);
    		HighGui.waitKey(0);

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 5][]

# 边缘检测 #

Imgproc.Canny(image, edges, threshold1, threshold2);

image 输入图像

edges 输出图像

threhold1 第一个阀值

threhold2 第一个阀值

# 边缘检测代码： #

String filename = "D:\\360MoveData\\Users\\lxn\\Desktop\\opencvtest\\lena.png";
    		Mat src = Imgcodecs.imread(filename);// 加载需要被蒙太奇的图片，原图
    		Mat dst = new Mat();
    		Imgproc.pyrUp(src, src);
    		//灰度化
    		Imgproc.cvtColor(src, dst, Imgproc.COLOR_BGR2GRAY);
    		//高斯滤波
    		Imgproc.GaussianBlur(dst, dst, new Size(3,3), 0.3);
    		Mat edage=new Mat();
    		
    		
    		Imgproc.Canny(dst,edage, 120,90,3);	
    
    		HighGui.imshow("原图", src);
    		HighGui.imshow("边缘检测并且绘制出边缘", edage);
    		HighGui.waitKey(0);

# 效果: #

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 6][]

**threshold1** first threshold for the hysteresis procedure.滞后程序的第一个阀值

**threshold2** second threshold for the hysteresis procedure. \\(=\\sqrt\{(dI/dx)^2 + (dI/dy)^2\}\\) should be used to calculate the image gradient magnitude ( L2gradient=true ), or whether the default \\(L\_1\\) norm \\(=|dI/dx|+|dI/dy|\\) is enough ( L2gradient=false ). 滞后程序的第二个阀值

Finds edges in an image using the Canny algorithm CITE: Canny86 . The function finds edges in the input image and marks them in the output map edges using the Canny algorithm. The smallest value between threshold1 and threshold2 is used for edge linking. The largest value is used to find initial segments of strong edges.

threshold1和threshold2之间的最小值用于边链接。较大值用于查找强边的初始段

上面是对imgeproc进行图片处理的几个方法，图片的放大放小，角点的检测，和边缘的检测

希望对你有所帮助

[20131227143130718]: /images/20221123/e11addda887d4f06a817a288c84c03be.png
[20200911153056234.png]: /images/20221123/c7613d460c074386a4ba1c867de092cf.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70]: /images/20221123/8053d2ce674345668e92cda878f5ebd2.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 1]: /images/20221123/907af5f2c8744f76b995e8c7ec8470e5.png
[20200911161237384.png]: /images/20221123/71df236abba64699a18f7be91144c466.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 2]: /images/20221123/7eb44cb9ee8a4183939fc364ff456e53.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 3]: /images/20221123/47fc7b9761324ab1a957fa12a14a85c1.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 4]: /images/20221123/ab1cc965be70438dad58acdefb5c0d6b.png
[20200911164140886.png]: /images/20221123/b1f83b6df1fd4d459f32206c46080fc1.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 5]: /images/20221123/483489641a7b429ab99d2ed6ce9b8b01.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2RhdG91bmlhbzE_size_16_color_FFFFFF_t_70 6]: /images/20221123/a21ea2624b5a47c683c607b645e57f61.png