OPENCV CONTOUR BASED MATCHING

I want to create a module with OPENCV like on below. But function of opencv which is cv.matchShapes(...), support only a contour for searching. How can I create a reference model which has multiple contours with OPENCV?

Information: (contour1 is not supported for multiple contours)

1. Python: cv2.matchShapes(<b>contour1</b>, contour2, method, parameter)

复制代码

Halcon Codes:

1. read_image (Test, 'test.jpg')
   
2. 
   
3. gen_rectangle1 (ROI_0, -3.64286, -31.5379, 152.201, 114.278)
   
4. reduce_domain(Test, ROI_0, ImageReduced)
   
5. 
   
6. create_shape_model (ImageReduced, 4, rad(0), rad(360), rad(1.6523), ['none','no_pregeneration'], 'use_polarity', [12,17,5], 4, ModelID)
   
7. get_shape_model_contours(ModelContours, ModelID, 1)
   
8. 
   
9. find_shape_model(Test, ModelID, -0.39, 0.79, 0.5, 1, 0.5, 'least_squares', 0, 0.9, Row, Column, Angle, Score)

复制代码

你可以自己扩展，加个循环，遍历比较多个轮廓。

for example, reference model has 5 contours and test image have 1000 contours. How can I match reference model with contours which is on image of test correctly?
According to me, It can be a mistake when contours are tried to matching one-to-one. Because, Test image can consist of more contours or less contours according to reference model.