Finding intersection point of incomplete circle with python

Question

I am trying to find the intersection point of the incomplete circle as image below:

Referring to this link's solution: Detect semi-circle in opencv

I am trying to convert the c++ code into python code, I have converted most of it but i don't understand the 2 lines of c++ code which are below:

1. Why does the radius need to divide by 25?

    // maximal distance of inlier might depend on the size of the circle float maxInlierDist = radius/25.0f;

2. I totally don't know how to convert this C++ line to python:

    if(dt.at<float>(cY,cX) < maxInlierDist)

I hope anyone could have help me on this, thank you!

I have tried to google around about some math formula but cannot find why radius need to divide by 25. I also not so good in c++.

My converted code:

# import the necessary packages
import numpy as np
import argparse
import cv2
import math

# construct the argument parser and parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-t", "--thres", required = True, help = "Path to the image")
ap.add_argument("-i", "--image", required = True, help = "Path to the image")
args = vars(ap.parse_args())

# load the image, clone it for output, and then convert it to grayscale
image = cv2.imread(args["image"])

gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

gray = cv2.Canny(gray, 200,20)

# detect circles in the image
circles = cv2.HoughCircles(gray, cv2.HOUGH_GRADIENT, 1,minDist=300,  
                               param1=200, param2=20,
                               minRadius=0, maxRadius=0)

#gray = (255*mask).astype(np.uint8)

dt = cv2.distanceTransform(255-gray, cv2.DIST_L2, 3)

cv2.imshow('Distance Transform', dt/255.0)
# ensure at least some circles were found

if circles is not None:
    # convert the (x, y) coordinates and radius of the circles to integers
    circles = np.round(circles[0, :]).astype("int")

    # loop over the (x, y) coordinates and radius of the circles
    for (x, y, r) in circles:
        # draw the circle in the output image, then draw a rectangle
        # corresponding to the center of the circle
        cv2.circle(image, (x, y), r, (0, 255, 0), 2)
        cv2.rectangle(image, (x - 5, y - 5), (x + 5, y + 5), 
                             (0, 128,255),-1)

        minInlierDist = 2.0
        counter =0
        inlier =0
        radius=r
        num_circle = 50

        maxInlierDist=radius/25.0
        if maxInlierDist<minInlierDist:
            maxInlierDist=minInlierDist

        for index in range(num_circle):
            counter +=1
            #angle = t * math.pi / 180
            angle = 2 * math.pi * index / num_circle

            cX = x + math.sin(angle)*radius
            cY = y + math.cos(angle)*radius
            centerxy = cX,cY
            cv2.circle(image,tuple(np.array(centerxy,int)),3,(0,0,255),-1)

#if(dt.at<float>(cY,cX) < maxInlierDist) #c++ ! I'm stuck here!

    cv2.imshow("output", image)#np.hstack([image, gray]))
    cv2.waitKey(0)
else:
    print("no circles found!")
    cv2.waitKey(0)

The most important question here is: How to know/indentify The green sampled dots on the circle are inliers and The blue dots are outliers.

Answer 1

1) 1/25 has been selected as an arbitrary fraction of the radius for the maximum acceptable error.

2) In opencv python, matrices are stored as multi-dimensional numpy arrays. To access the point at (cY,cX), use dt[cY,cX]