(12) findContours function hierarchy explanation

欢迎访问个人网络日志知行空间

1.基本用法

Get the outline of the object,Generally, it is best to first grayscale the image and then perform thresholding,and then used to detect contours.

void cv::findContours(
     InputOutputArray image,
     OutputArrayOfArrays contours,
     OutputArray hierarchy,
     int mode,
     int method,
     Point offset = Point() 
)

• image: 8位单通道图像
• contours: 检测到的轮廓,vector<vector<cv::Point>>
• hierarchy: Nesting and proximity relationships between detected contours,vector<cv::Vec4i>
• mode: Different contour detection algorithms,常用的有RETR_EXTERNAL/RETR_LIST/RETR_CCOMP/RETR_TREE四种
• method: 轮廓逼近方法,见,Contours can be represented using fewer points,减少内存占用.
• offset: The amount by which the contour points should be offset,当在roiIt is useful when you want to restore the original image after extracting the contour.

2.轮廓提取模式

第四个参数mode,Different contour extraction algorithms can be selected,常用的有RETR_EXTERNAL/RETR_LIST/RETR_CCOMP/RETR_TREE四种.下面分别进行介绍.在findContours函数中,其第3个参数hierarchyRepresents a hierarchical relationship between contours,对于不同modecontour extraction algorithm,其返回的值是不同的.如下图【来自于OpenCV Doc】：

Different contours in the graph have hierarchical embedded relationships,We call the outer contours 父,The inner contour is called子,hierarchyIt is a matrix representing the parent-child and adjacent relationship of contours.上图中有0/1/2/3/4/5/2a/3a 8 个轮廓,其中0,1,2is the outermost contour,Can be recorded as they are in the hierarchy0hierarchy-0.而2a是轮廓2的子轮廓,Note that it is in the hierarchy1hierarchy 1.同样轮廓3是轮廓2a的子轮廓,Note that it is in the hierarchy2hierarchy 2.同样轮廓3a是轮廓3的子轮廓,Note that it is in the hierarchy3hierarchy 3.4/5是3a的子轮廓,its composition hierarchy4hierarchy 4. The contours belonging to each layer have their own information,Such as what its sub-contours are,What is the parent profile,OpenCVThe relationship of each contour to other contours is represented by a four-element array,The values ​​in this four-dimensional array are represented separately**[Next, Previous, First_Child, Parent], NextIndicates that they belong to the same levelhierarchythe next contour,Outline above0为例,0,1,2belong to the same levelhierarchy-0的轮廓,因此0的Next是1,1的Next是2.contours in the same level2已经是最后一个了,因此其Next是-1.The same goes for the contours in the image above4,It belongs to the same level4hierarchy-4的轮廓是5,因此4的Next=5,而5的Next=-1.PreviousRepresents the previous contour in the same level**,如上图,1的Previous=0, 2的Previous=1,0的Previous=-1.First_ChildRepresents the first subcontour of the current contour,如上图,0的First_Child=-1,2的First_Child=2a,3a的First_Child=4.ParentRepresents the parent contour of the current contour,如上图,4和5The parent contours of 3a,3a的父轮廓是3,3的父轮廓是2a,2的Parent=-1.

findContours方法中的modeparameters will return differenthierarchy信息,Because some algorithms will find the nesting and adjacent relationship between contours,Some just find the contours without parsing the information between the contours.

2.1 RETR_LIST

RETR_LISTThe algorithm will only return contour information,There is no nesting information between contours.因此,All contours belong to the same levelhierarchy没有父子关系, hierarchy返回值中只有Next和Previous,Parent和First_Child都为-1.The first of a four-dimensional array3和第4个元素都是-1.Run as shown abovefindContours后的输出：

findContours(image, contours, hierarchy, cv::RETR_LIST, cv::CHAIN_APPROX_SIMPLE);

>>> hierarchy
[1, -1, -1, -1]
[2, 0, -1, -1]
[3, 1, -1, -1]
[4, 2, -1, -1]
[5, 3, -1, -1]
[6, 4, -1, -1]
[7, 5, -1, -1]
[-1, 6, -1, -1]

这里的0/1/2/3/4/5/6/7Correspondingly, the outline is incontours中的下标.

2.2 RETR_EXTERNAL

RETR_EXTERNAL算法,Only the outermost contour information will be returned,All subcontours are not returned,如上图,使用RETR_EXTERNALThe algorithm will just returnhierarchy-0层级0的3个轮廓.当然hierarchy中也只有3proximity information between contours,Parent和First_Child依然都为-1.

findContours(image, contours, hierarchy, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_SIMPLE);

[1, -1, -1, -1]
[2, 0, -1, -1]
[-1, 1, -1, -1]

2.3 RETR_CCOMP

RETR_CCOMPThe algorithm will find all the contours in the graph,But only the outlines will be organized into two layershierarchy=2.The outer contour of the object belongs tohierarchy-0,The inner contour belongs tohierarchy-1,如上图0/1/2/3/4/5都属于hierarchy-0,而2a/3a属于hierarchy-1.

findContours(image, contours, hierarchy, cv::RETR_CCOMP, cv::CHAIN_APPROX_SIMPLE);

[1, -1, -1, -1]
[2, 0, -1, -1]
[4, 1, 3, -1]
[-1, -1, -1, 2]
[6, 2, 5, -1]
[-1, -1, -1, 4]
[7, 4, -1, -1]
[-1, 6, -1, -1]

It's worth noting for contours2a和3a,其hierarchy分别是[-1, -1, -1, 2]和[-1, -1, -1, 4].这是因为2a和3a虽然属于hierarchy-1,But there is still an outline in the middle3,因此2a和3aThere is no proximity relationship between them.

再来看个例子,如下图：【来自于OpenCV Doc】

轮廓0is the outer contour,1和2respectively the outline0The inner contour of the enclosed object,4belongs to the inner contour,3belong to the outer contour,6belongs to the inner contour,5belong to the outer contour,7和8Also belong to the outer contour.对于轮廓0其属于hierarchy-1,Two inner contours1和2属于hierarchy-2.So for contours0,其Next=3,same levelhierarchy level的下一个,previous=-1,‵First-Child=1,so outline0的hierarchy=[3,-1, 1, -1]`.

轮廓1belong to the hierarchy2,hierarchy-2,its next at the same level(与1in the same parent outline)轮廓是2,其他均为-1,因此轮廓1的hierarchy=[2, -1, -1, 0].

轮廓2belong to the hierarchy2,hierarchy-2,Its previous contour is under the same parent outer contour1,其余为-1,因此轮廓2的hierarchy=[-1, 1, -1, 0].

轮廓3在hierarchy-1中的Next=5,Previous=0,First-Child=4,Parent=-1.

>>> hierarchy
array([[[ 3, -1,  1, -1],
        [ 2, -1, -1,  0],
        [-1,  1, -1,  0],
        [ 5,  0,  4, -1],
        [-1, -1, -1,  3],
        [ 7,  3,  6, -1],
        [-1, -1, -1,  5],
        [ 8,  5, -1, -1],
        [-1,  7, -1, -1]]])

2.4 RETR_TREE

RETR_TREEThe algorithm extracts all contours,and returns the nested relationship between all contours.如上图,使用RETR_TREEthe resulting contourhierarchy之间的关系为：【来自于OpenCV Doc】The green words in parentheses indicate the level to which the contour belongshierarchy.

以轮廓0为例,其属于hierarchy-0,‵Next=7, Previous=-1, First_Child=1, Parent=-1`.

>>> hierarchy
array([[[ 7, -1,  1, -1],
        [-1, -1,  2,  0],
        [-1, -1,  3,  1],
        [-1, -1,  4,  2],
        [-1, -1,  5,  3],
        [ 6, -1, -1,  4],
        [-1,  5, -1,  4],
        [ 8,  0, -1, -1],
        [-1,  7, -1, -1]]])

findContours(image, contours, hierarchy, cv::RETR_CCOMP, cv::CHAIN_APPROX_SIMPLE);

[3, -1, 1, -1]
[2, -1, -1, 0]
[-1, 1, -1, 0]
[5, 0, 4, -1]
[-1, -1, -1, 3]
[6, 3, -1, -1]
[-1, 5, -1, -1]

3.测试代码

#include <opencv2/opencv.hpp>
#include <common.h>

using namespace std;

int main(int argc, char **argv) 
{
    
    cv::RNG rng(12345);
    cout << "Usage: " << argv[0] << "\n";
    cv::String img_dir = "/images/OpenCV/2findContours/hole-hierarchy.png";
    cv::Mat image = cv::imread(img_dir, cv::IMREAD_GRAYSCALE);

    vector<cv::Vec4i> hierarchy;
    vector<vector<cv::Point> > contours;

    findContours(image, contours, hierarchy, cv::RETR_CCOMP, cv::CHAIN_APPROX_SIMPLE);
    cv::cvtColor(image, image, cv::COLOR_GRAY2BGR);
    std::cout << "Contours Size: " << contours.size() << std::endl;
    for(size_t i=0; i<contours.size(); i++)
    {
    
        cv::Scalar clr = cv::Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255));
        cv::drawContours(image, contours, i, clr, 3);
    }

    for(auto &v : hierarchy)
    {
    
        std::cout << v << std::endl;
    }
    std::cout << image.channels() << std::endl;
    cv::imwrite("contours.png", image);
    cv::imshow("Img", image);
    cv::waitKey(0);
    return 0;
}s

参考资料

• 1.https://docs.opencv.org/4.x/d9/d8b/tutorial_py_contours_hierarchy.html

欢迎访问个人网络日志知行空间