个人中心
gpt4 book ai didi

使用OpenCV去除面积较小的连通域

转载 作者：qq735679552 更新时间：2022-09-29 22:32:09 27 4
gpt4 key购买 nike

CFSDN坚持开源创造价值，我们致力于搭建一个资源共享平台，让每一个IT人在这里找到属于你的精彩世界.

这篇CFSDN的博客文章使用OpenCV去除面积较小的连通域由作者收集整理，如果你对这篇文章有兴趣，记得点赞哟.

这是后期补充的部分，和前期的代码不太一样 。

效果图 。

使用OpenCV去除面积较小的连通域

源代码 。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
/ / 测试
void CCutImageVS2013Dlg::OnBnClickedTestButton1()
{
     vector<vector<Point> > contours;  / / 轮廓数组
     vector<Point2d> centers;    / / 轮廓质心坐标
     vector<vector<Point> >::iterator itr; / / 轮廓迭代器
     vector<Point2d>::iterator itrc;  / / 质心坐标迭代器
     vector<vector<Point> > con;   / / 当前轮廓
 
     double area;
     double minarea = 1000 ;
     double maxarea = 0 ;
     Moments mom;       / / 轮廓矩
     Mat image, gray, edge, dst;
     image = imread( "D:\\66.png" );
     cvtColor(image, gray, COLOR_BGR2GRAY);
     Mat rgbImg(gray.size(), CV_8UC3); / / 创建三通道图
     blur(gray, edge, Size( 3 , 3 ));       / / 模糊去噪
     threshold(edge, edge, 200 , 255 , THRESH_BINARY_INV); / / 二值化处理,黑底白字
     / / - - - - - - - - 去除较小轮廓，并寻找最大轮廓 - - - - - - - - - - - - - - - - - - - - - - - - - -
     findContours(edge, contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_SIMPLE); / / 寻找轮廓
     itr = contours.begin();    / / 使用迭代器去除噪声轮廓
     while (itr ! = contours.end())
     {
         area = contourArea( * itr);  / / 获得轮廓面积
         if (area<minarea)    / / 删除较小面积的轮廓
         {
             itr = contours.erase(itr); / / itr一旦erase，需要重新赋值
         }
         else
         {
             itr + + ;
         }
         if (area>maxarea)    / / 寻找最大轮廓
         {
             maxarea = area;
         }
     }
     dst = Mat::zeros(image.rows, image.cols, CV_8UC3);
     / * 绘制连通区域轮廓，计算质心坐标 * /
     Point2d center;
     itr = contours.begin();
     while (itr ! = contours.end())
     {
         area = contourArea( * itr);      
         con.push_back( * itr);   / / 获取当前轮廓
         if (area = = maxarea)
         {
             vector<Rect> boundRect( 1 ); / / 定义外接矩形集合
             boundRect[ 0 ] = boundingRect(Mat( * itr));
             cvtColor(gray, rgbImg, COLOR_GRAY2BGR);
             Rect select;
             select.x = boundRect[ 0 ].x;
             select.y = boundRect[ 0 ].y;
             select.width = boundRect[ 0 ].width;
             select.height = boundRect[ 0 ].height;
             rectangle(rgbImg, select, Scalar( 0 , 255 , 0 ), 3 , 2 ); / / 用矩形画矩形窗
             drawContours(dst, con, - 1 , Scalar( 0 , 0 , 255 ), 2 ); / / 最大面积红色绘制
         }
         else
             drawContours(dst, con, - 1 , Scalar( 255 , 0 , 0 ), 2 ); / / 其它面积蓝色绘制
         con.pop_back();
         / / 计算质心
         mom = moments( * itr);
         center.x = ( int )(mom.m10 / mom.m00);
         center.y = ( int )(mom.m01 / mom.m00);
         centers.push_back(center);
         itr + + ;
     }
     imshow( "rgbImg" , rgbImg);
     / / imshow( "gray" , gray);
     / / imshow( "edge" , edge);
     imshow( "origin" , image);
     imshow( "connected_region" , dst);
     waitKey( 0 );
     return ;
}

前期做的，方法可能不太一样 。

一，先看效果图 。

原图 。

使用OpenCV去除面积较小的连通域

处理前后图 。

使用OpenCV去除面积较小的连通域

二，实现源代码 。

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
/ / = = = = = = = 函数实现 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
void RemoveSmallRegion(Mat &Src, Mat &Dst, int AreaLimit, int CheckMode, int NeihborMode)
{
     int RemoveCount = 0 ;
     / / 新建一幅标签图像初始化为 0 像素点，为了记录每个像素点检验状态的标签， 0 代表未检查， 1 代表正在检查, 2 代表检查不合格（需要反转颜色）， 3 代表检查合格或不需检查
     / / 初始化的图像全部为 0 ，未检查
     Mat PointLabel = Mat::zeros(Src.size(), CV_8UC1);
     if (CheckMode = = 1 ) / / 去除小连通区域的白色点
     {
         / / cout << "去除小连通域." ;
         for ( int i = 0 ; i < Src.rows; i + + )
         {
             for ( int j = 0 ; j < Src.cols; j + + )
             {
                 if (Src.at<uchar>(i, j) < 10 )
                 {
                     PointLabel.at<uchar>(i, j) = 3 ; / / 将背景黑色点标记为合格，像素为 3
                 }
             }
         }
     }
     else / / 去除孔洞，黑色点像素
     {
         / / cout << "去除孔洞" ;
         for ( int i = 0 ; i < Src.rows; i + + )
         {
             for ( int j = 0 ; j < Src.cols; j + + )
             {
                 if (Src.at<uchar>(i, j) > 10 )
                 {
                     PointLabel.at<uchar>(i, j) = 3 ; / / 如果原图是白色区域，标记为合格，像素为 3
                 }
             }
         }
     }
     vector<Point2i>NeihborPos; / / 将邻域压进容器
     NeihborPos.push_back(Point2i( - 1 , 0 ));
     NeihborPos.push_back(Point2i( 1 , 0 ));
     NeihborPos.push_back(Point2i( 0 , - 1 ));
     NeihborPos.push_back(Point2i( 0 , 1 ));
     if (NeihborMode = = 1 )
     {
         / / cout << "Neighbor mode: 8邻域." << endl;
         NeihborPos.push_back(Point2i( - 1 , - 1 ));
         NeihborPos.push_back(Point2i( - 1 , 1 ));
         NeihborPos.push_back(Point2i( 1 , - 1 ));
         NeihborPos.push_back(Point2i( 1 , 1 ));
     }
     else int a = 0 ; / / cout << "Neighbor mode: 4邻域." << endl;
     int NeihborCount = 4 + 4 * NeihborMode;
     int CurrX = 0 , CurrY = 0 ;
     / / 开始检测
     for ( int i = 0 ; i < Src.rows; i + + )
     {
         for ( int j = 0 ; j < Src.cols; j + + )
         {
             if (PointLabel.at<uchar>(i, j) = = 0 ) / / 标签图像像素点为 0 ，表示还未检查的不合格点
             { / / 开始检查
                 vector<Point2i>GrowBuffer; / / 记录检查像素点的个数
                 GrowBuffer.push_back(Point2i(j, i));
                 PointLabel.at<uchar>(i, j) = 1 ; / / 标记为正在检查
                 int CheckResult = 0 ;
                 for ( int z = 0 ; z < GrowBuffer.size(); z + + )
                 {
                     for ( int q = 0 ; q < NeihborCount; q + + )
                     {
                         CurrX = GrowBuffer.at(z).x + NeihborPos.at(q).x;
                         CurrY = GrowBuffer.at(z).y + NeihborPos.at(q).y;
                         if (CurrX > = 0 && CurrX<Src.cols&&CurrY > = 0 && CurrY<Src.rows) / / 防止越界
                         {
                             if (PointLabel.at<uchar>(CurrY, CurrX) = = 0 )
                             {
                                 GrowBuffer.push_back(Point2i(CurrX, CurrY)); / / 邻域点加入 buffer
                                 PointLabel.at<uchar>(CurrY, CurrX) = 1 ;   / / 更新邻域点的检查标签，避免重复检查
                             }
                         }
                     }
                 }
                 if (GrowBuffer.size()>AreaLimit) / / 判断结果（是否超出限定的大小）， 1 为未超出， 2 为超出
                     CheckResult = 2 ;
                 else
                 {
                     CheckResult = 1 ;
                     RemoveCount + + ; / / 记录有多少区域被去除
                 }
                 for ( int z = 0 ; z < GrowBuffer.size(); z + + )
                 {
                     CurrX = GrowBuffer.at(z).x;
                     CurrY = GrowBuffer.at(z).y;
                     PointLabel.at<uchar>(CurrY, CurrX) + = CheckResult; / / 标记不合格的像素点，像素值为 2
                 }
                 / / * * * * * * * * 结束该点处的检查 * * * * * * * * * *
             }
         }
     }
     CheckMode = 255 * ( 1 - CheckMode);
     / / 开始反转面积过小的区域
     for ( int i = 0 ; i < Src.rows; + + i)
     {
         for ( int j = 0 ; j < Src.cols; + + j)
         {
             if (PointLabel.at<uchar>(i, j) = = 2 )
             {
                 Dst.at<uchar>(i, j) = CheckMode;
             }
             else if (PointLabel.at<uchar>(i, j) = = 3 )
             {
                 Dst.at<uchar>(i, j) = Src.at<uchar>(i, j);
             }
         }
     }
     / / cout << RemoveCount << " objects removed." << endl;
}
/ / = = = = = = = 函数实现 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
/ / = = = = = = = 调用函数 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =
     Mat img;
     img = imread( "D:\\1_1.jpg" , 0 ); / / 读取图片
     threshold(img, img, 128 , 255 , CV_THRESH_BINARY_INV);
     imshow( "去除前" , img);
     Mat img1;
     RemoveSmallRegion(img, img, 200 , 0 , 1 );
     imshow( "去除后" , img);
     waitKey( 0 );
/ / = = = = = = = 调用函数 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =

以上这篇使用OpenCV去除面积较小的连通域就是小编分享给大家的全部内容了，希望能给大家一个参考，也希望大家多多支持我.

原文链接：https://blog.csdn.net/sxlsxl119/article/details/80493655 。

最后此篇关于使用OpenCV去除面积较小的连通域的文章就讲到这里了,如果你想了解更多关于使用OpenCV去除面积较小的连通域的内容请搜索CFSDN的文章或继续浏览相关文章，希望大家以后支持我的博客！ 。

27 4 0
文章推荐： Python OpenCV去除字母后面的杂线操作
文章推荐： Redis和springboot 整合redisUtil类的示例代码
文章推荐： 学python对电脑配置有要求么
文章推荐： 浅谈OpenCV中的新函数connectedComponentsWithStats用法
qq735679552
个人简介

我是一名优秀的程序员,十分优秀！

滴滴打车优惠券免费领取
滴滴打车优惠券
Copyright 2021 - 2024 cfsdn All Rights Reserved 蜀ICP备2022000587号
广告合作:1813099741@qq.com 6ren.com