Python OpenCV轮廓排序

Question

I am working with words scanned from old text. I want to be able to read the letters from left to right. Unfortunately, when I try to sort the contours based based on the cv2.boundingRect x value it doesn't sort that way I'd like. I've included an image numbering the letters on the order in which they are sorted. The outcome I wanted was for dead to be 0,1,2,3 and battery 4, 5, 6, 7 essentially reading dead then battery.

Here is my code:

def sort_contours(image, area_threshold):
    contours, hierarchy = cv2.findContours(image, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
    all_contours = []
    for contour in contours:
        area = cv2.contourArea(contour)
        x, y, w, h = cv2.boundingRect(contour)

        if area > area_threshold:
            all_contours.append((x, y, w, h))
    return sorted(all_contours, key=lambda tup: tup[0])

Answer 1

You need to compute the distance from the origin (0,0) to the starting point. I have tried in a dummy image which I created and it works.

You need to update your sorting algorithm with the below code. which computes the Euclidean distance from the origin.

im2, contours, hierarchy = cv2.findContours(thresh,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE)

points = []
for cnt in contours:
    x,y,w,h=cv2.boundingRect(cnt)

    points.append((x,y,w,h))

#print(points)
points = sorted(points,key=lambda data:math.sqrt(data[0]**2+data[1]**2));

The below image is the output.