CODEWITHSUNDEEP
pythonopencv
Susie B
Susie B

Reputation: 3

loop for image processing in python / attribute error 'NoneType'

I have almost 900 .jpg images (obtained by splitting a video in frames, each image is 116ko), and I want to compare each of my frames to the first one (= frame0) using the "Mean Squared Error".

It is quite easy to load two images and compare them, but it is much more difficult (for me at least) to write a loop that loads and compares my images. My aim is just to have a list with all the "Mean Squared Error" (MSE) values in it.

Issues:

I don't need to have all the images loaded or opened at the same time, but I don't know how to load an image just during the processing time.

My images are in the same folder as my .py file.

When I run my code, it returns:

Traceback (most recent call last): File "C:\Users\Susie\Documents\programmes_python\python\images_video\comparaison-images2.py", line 46, in compare_images(frame0, frame0) File "C:\Users\Susie\Documents\programmes_python\python\images_video\comparaison-images2.py", line 24, in compare_images s = ssim(imageA, imageB) File "C:\Python27\lib\site-packages\skimage\measure_structural_similarity.py", line 151, in compare_ssim "win_size exceeds image extent. If the input is a multichannel " ValueError: win_size exceeds image extent. If the input is a multichannel (color) image, set multichannel=True.

# import the necessary packages
from skimage.measure import compare_ssim as ssim
import numpy as np
import cv2
from os import listdir
from os.path import isfile, join

def mse(imageA, imageB):
    # the 'Mean Squared Error' between the two images is the
    # sum of the squared difference between the two images
    err = np.sum((imageA.astype("float") - imageB.astype("float")) ** 2)
    err /= float(imageA.shape[0] * imageA.shape[1])

    # return the MSE, the lower the error, the more "similar"
    # the two images are
    return err

def compare_images(imageA, imageB):
    # compute the mean squared error and structural similarity
    # index for the images
    m = mse(imageA, imageB)
    s = ssim(imageA, imageB)
    mse_list.append(m)
    ssim_list.append(s)
    print (mse_list)
    print (ssim_list)


# load the images and convert the images to grayscale
frame0 = cv2.imread("frame0.jpg")

mypath='C:\Users\Susie\Documents\programmes_python\python\images_video'
onlyfiles = [ f for f in listdir(mypath) if isfile(join(mypath,f)) ]
images = np.empty(len(onlyfiles), dtype=object)

for n in range(1, len(onlyfiles)):
  images[n] = cv2.imread( join(mypath,onlyfiles[n]) )
  images[n] = cv2.cvtColor(images[n], cv2.COLOR_BGR2GRAY)


# compare the images
mse_list = []
ssim_list = []
compare_images(frame0, frame0)

for n in range(1, len(onlyfiles)):
    compare_images(frame0, images[n])

Upvotes: 0

Views: 2198

Answers (2)

micmalti
micmalti

Reputation: 571

Since you're using grayscale images, you need to specify this by adding the 0 parameter as follows:

frame0 = cv2.imread("frame0.jpg", 0)

Check OpenCV documentation for more detail: https://docs.opencv.org/3.0-beta/modules/imgcodecs/doc/reading_and_writing_images.html

Upvotes: 0

Jeru Luke
Jeru Luke

Reputation: 21223

I would like to mention a couple of programming mistakes:

1. Attribute error :

Ensure that the .py file and the collection of images are present within the same folder. If they are not, mention the path exclusively for eg. mypath='C:/Users/Susie/Documents/programmes_python/python/images_video

2. Value error :

Comparing a multi-channel image with a single channel image throws this error. Make sure the images are of the same size and have same number of channels.

Making mistakes is a step to learning !!!

Upvotes: 1

Related Questions