shifted image similarity feature

.py

image_a = [
  [1,0,0,0,1,0,1],
  [0,1,0,1,0,0,1],
  [0,1,0,0,0,1,1],
  [0,0,1,1,1,0,0],
  [0,0,0,0,1,1,1],
  [0,1,0,0,0,1,1],
  [0,0,0,1,1,1,0],
]
image_b = [
  [0,0,0,1,0,1,1],
  [1,0,1,0,0,1,0],
  [1,0,0,0,1,1,0],
  [0,1,1,1,0,0,0],
  [0,0,0,1,1,1,0],
  [1,0,0,0,1,1,0],
  [0,0,1,1,1,0,0],
]

def chunkify(image, chunk_size=3):
  border = chunk_size//2
  for i in range(border, len(image)-border):
    for j in range(border, len(image[i])-border):
      #print(j, border)
      yield [
        image[i-border][j-border:j+border+1],
        image[border][j-border:j+border+1],
        image[i+border][j-border:j+border+1]
      ]

def compare_images(image, other):
  chunks = [*chunkify(image)]  # this should be a counter-dict
  total_count = len(chunks)
  same_count = 0
  for chunk in chunkify(other):
    try:
      chunks.remove(chunk)
      same_count += 1
    except ValueError:
      pass
  return same_count / total_count

print(compare_images(image_a, image_b))  # 0.84