select and seperate wall with canny

Load your image and convert it to grayscale, as edge detection works on single-channel images.   

   import cv2
   import numpy as np
   import matplotlib.pyplot as plt

   image = cv2.imread('image.png')
   gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

Then use the Canny algorithm to detect edges in the image.

   edges = cv2.Canny(gray, threshold1=100, threshold2=200)

Find contours from the edges detected. These contours will help you identify separate walls.

   contours, _ = cv2.findContours(edges, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)

 Draw each contour and fill them with a color.

   filled_image = np.zeros_like(image)

   for contour in contours:
       cv2.drawContours(filled_image, [contour], -1, (255, 255, 255), thickness=cv2.FILLED)

Then display the original image, edges, and the filled contours.

   plt.figure(figsize=(10, 5))

   plt.subplot(1, 3, 1)
   plt.title('Original Image')
   plt.imshow(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))

   plt.subplot(1, 3, 2)
   plt.title('Canny Edges')
   plt.imshow(edges, cmap='gray')

   plt.subplot(1, 3, 3)
   plt.title('Filled Walls')
   plt.imshow(cv2.cvtColor(filled_image, cv2.COLOR_BGR2RGB))

   plt.show()