Why Does Hough Transform Only Produce One Lines When There's Multiple Lines?-CodePudding

I'm attempting to use hough transform to find the lines that I need. I'm working with multiple images and an example of one is this

https://i.stack.imgur.com/KmGdP.jpg

However when using the hough transform the result comes out as

https://i.stack.imgur.com/DJvbg.jpg

I don't know why only one line ends up on the sign in which I want the hough lines to encompass all four sides, and with the other images I have some don't even have hough lines showing up on them. What am i doing wrong? Is there any corrections to be put in place? This is the code I'm running

import numpy as np
import cv2 as cv
import os

images = []
edges = []

light_range = (0, 0, 0)
dark_range = (80, 80, 80)

#Importing testing images
for i in os.listdir("Directional Signage/Train"):
    img = cv.imread(os.path.join("Directional Signage/Train", i))

    if img is None:
        print("Couldn't read this image"   str(i))
    else:
        images.append(img)


for i in range(0, len(images)):
    #Preprocessing 
    #mask = cv.inRange(images[i], light_range, dark_range)
    img = np.float32(images[i]) / 255.0

    gx = cv.Sobel(img, cv.CV_32F, 1, 0, ksize=1)
    gy = cv.Sobel(img, cv.CV_32F, 0, 1, ksize=1)

    mag, angle = cv.cartToPolar(gx, gy, angleInDegrees=True)
    gray = cv.cvtColor(mag,cv.COLOR_BGR2GRAY)
    gray = np.uint8(gray * 255.0)
    edges = cv.Canny(gray,50,150,apertureSize = 3)
    lines = cv.HoughLines(edges,2,np.pi/90,200)
    #edges.append(lines)

    for rho,theta in lines[0]:
        a = np.cos(theta)
        b = np.sin(theta)
        x0 = a*rho
        y0 = b*rho
        x1 = int(x0   1000*(-b))
        y1 = int(y0   1000*(a))
        x2 = int(x0 - 1000*(-b))
        y2 = int(y0 - 1000*(a))

        cv.line(images[i],(x1,y1),(x2,y2),(0,0,255),2)


for i in range(0, len(images)):
    cv.imshow("blobby "   str(i), images[i])

l = cv.waitKey(0)

CodePudding user response：

You're just iterating over one line (lines[0]), so it is obvious that you get one line from each image, do this instead:

[...]
    for line in lines:
        for rho,theta in line:
            a = np.cos(theta)
            b = np.sin(theta)
            x0 = a*rho
            y0 = b*rho
            x1 = int(x0   1000*(-b))
            y1 = int(y0   1000*(a))
            x2 = int(x0 - 1000*(-b))
            y2 = int(y0 - 1000*(a))

            cv.line(images[i],(x1,y1),(x2,y2),(0,0,255),2)
[...]

Update: now rather than getting a single red line my entire image goes red
The reason is that, the Hough function finds every lines but by priority. The lines which are more likely to be a real line are at first part of the list.
First two lines of the list, maybe is what you're looking for, so try this:

[...]
    for line in lines[:2]:
        for rho,theta in line:
            a = np.cos(theta)
            b = np.sin(theta)
            x0 = a*rho
            y0 = b*rho
            x1 = int(x0   1000*(-b))
            y1 = int(y0   1000*(a))
            x2 = int(x0 - 1000*(-b))
            y2 = int(y0 - 1000*(a))

            cv.line(images[i],(x1,y1),(x2,y2),(0,0,255),2)
[...]