computer vision/ defect detection/ python

Question

I want to get the row of the red box abnormal object, I divide the rows seperately, and then use the first object as the base, mse and ssim methods are adopted to compare the rest targets of each row with the defined base object, aims to find the abnormal objects like the reb box.

the main method as follow: imgs is a list contains the seperated object of a row

def compare_images(imageA, imageB,title):
    # compute the mean squared error and structural similarity
    # index for the images
    m = mse(imageA, imageB)
    # print 'The Calculate Time of Method MSE is: %.5f' %(end-start)
    s = ssim(imageA, imageB)
    # print 'The Calculate Time of Method SSIM is: %.5f' %(end-start)
    return m, s

threshold = 0
for i in range(len(imgs)):
    # cv2.namedWindow('img', cv2.WINDOW_NORMAL)
    # cv2.imshow('img', imgs[i])
    # cv2.waitKey()
    [c, d] = compare_images(base, imgs[i], "Original vs. contrast")
    if d>0.6:
       base = imgs[i].copy()
    if d<0.1:
       threshold+=1
    if threshold>5:
       break
    print("MSE:{}, SSIM:{}".format(c, d))