[FIXED] Is there any way to create multiple bounding boxes around non zero numpy array values?

Issue

I have a numpy array of an image. I have set non zero values for my region of interest and for rest I have set value as 0. Now if I were to create a single bounding box, I could check for 1st occurance and last occurance of non zero value and get the coordinates. But what if the non zero values are on different places?

How can I create 2 bounding boxes instead of one?

I tried -

A = array([[0, 0, 0, 0, 0, 0, 0],
           [1, 1, 0, 0, 0, 0, 0],
           [1, 1, 0, 0, 0, 0, 0],
           [0, 0, 0, 0, 1, 1, 1],
           [0, 0, 0, 0, 1, 1, 1],
           [0, 0, 0, 0, 1, 1, 1],
           [0, 0, 0, 0, 0, 0, 0]])

and I was expecting 2 bounding boxes around the Ones.

Solution

I tested similar code in my own project, but I didn't test this particular snippet. Comment if you encounter any bugs.

import cv2
import numpy as np

A = np.array([[0, 0, 0, 0, 0, 0, 0],
           [1, 1, 0, 0, 0, 0, 0],
           [1, 1, 0, 0, 0, 0, 0],
           [0, 0, 0, 0, 1, 1, 1],
           [0, 0, 0, 0, 1, 1, 1],
           [0, 0, 0, 0, 1, 1, 1],
           [0, 0, 0, 0, 0, 0, 0]])

_, markers = cv2.connectedComponents(np.expand_dims(A.astype(np.uint8),axis=2),connectivity=4)

markers = np.squeeze(markers)

bounding_boxes = []

for label in range(1,int(np.max(markers))+1):

    locations = np.transpose(np.nonzero(markers==label))

    min_x, max_x, min_y, max_y = np.min(locations[:,1]), np.max(locations[:,1]), np.min(locations[:,0]), np.max(locations[:,0])

    bounding_boxes.append((min_x, max_x, min_y, max_y))

print(bounding_boxes)

Answered By - Michael Sohnen