Stack multiple images in python pillow -

i trying image stacking in python pillow. take large number of images (say 10), , each pixel, take median value this: http://petapixel.com/2013/05/29/a-look-at-reducing-noise-in-photographs-using-median-blending/.

right now, can in incredibly brute force manner (using getpixel , put pixel), takes long time.

here have far:

import os pil import image files = os.listdir("./") new_im = image.new('rgb', (4000,3000)) ims={} in range(10,100):     ims[i]={}                im=image.open("./"+files[i])     x in range(400):         ims[i][x]={}                     y in range(300):             ims[i][x][y]=im.getpixel((x,y))  x in range(400):     y in range(300):         these1=[]         these2=[]         these3=[]         in ims:             these1.append(ims[i][x][y][0])             these2.append(ims[i][x][y][1])             these3.append(ims[i][x][y][2])         these1.sort()         these2.sort()         these3.sort()         new_im.putpixel((x,y),(these1[len(these1)/2],these2[len(these2)/2],these3[len(these3)/2]))  new_im.show() 

you can vectorise lot of these loops arrays. example np.array(im) return array of pixels, shape (400, 300, 3). store in array.

image_stacks = np.zeros(shape=(10, 400, 300, 3), dtype=np.uint8) in xrange(image_stacks.shape[0]):     # open image file , store in variable `im`,     image_stacks[i] = np.array(im) 

now can calculate median preferred way, numpy has method that, too.

image_median = np.median(image_stacks, axis=0).astype(np.uint8) image = image.fromarray(image_median) 

two things notice here np.median() return float type array, , want convert unsigned int8. other thing if number of elements even, median calculated mean of 2 middle values, may end being odd number divided two, such 13.5. when converted integer, rounded down. such minor precision loss should not visually affect result.