How about we open a image, do some calcs and save it?
This is the first post about it, so how can we open a image in python? usually images can be in a lot of different formats so to solve this puzze and to simplify the thinks a little bit we will use a tool called PIL (Python Imaging Library).
To simplify the things we will use a library called NumPy (Numeric Python) to deal with matrices.
First of all, if you don’t have PIL or Numpy install it. On ubuntu you can just type “sudo apt-get install python-imaging python-numpy”
File: threshold.py
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 | #!/usr/bin/python #-*- coding: utf-8 -*- # # Author: Fernando Paolieri Neto | 07-aug-2011 # #import needed libraries import Image import numpy import sys #function to convert numpy to pil def array2pil(arr): nd = len(arr.shape) x = arr.astype('B') if nd == 2: d, h, w = (1,) + arr.shape mode = 'L' elif nd == 3: if arr.dtype.char == '?': raise TypeError, "Binary array cannot be RGB" h, w, d = arr.shape if d == 1: mode = 'L' elif d == 3: mode = 'RGB' elif d == 4: mode = 'RGBA' else: raise TypeError, "Array first dimension must be 1, 3 or 4 (%d)" % d else: raise TypeError, "Array must have 2 or 3 dimensions (%d)" % nd pil = Image.fromstring(mode, (w,h), x.tostring()) if arr.dtype.char == '?': pil = pil.point(lambda i: i>0, '1') return pil if __name__ == "__main__": #read image using Pil img = Image.open(sys.argv[1]) #convert Pil image to Numpy Array F = numpy.asarray(img) #if -B is passed by argument we will do the averege #of the 3 channels R,G and B if '-B' in sys.argv: F= (F[:,:,0]/3 + F[:,:,1]/3 + F[:,:,2]/3) #Do the threshold F = (F>127)*numpy.uint8(255) #If -I is passed, the image will be inverted if '-I' in sys.argv: F = 255-F #convert to pil img = array2pil(F) #save output img.save("out.png","PNG") |
Save your file as threshold.py or get it from GitRepository.
Now lets us just run the program (example on ubuntu):
python threshold.py imageIn.jpg -B -I
If you set -B will be used the average, and if you set -I the image will be inverted.
As result a new image called out.png will be generated.
A demo was build in CGI to show this code running on a webserver. Threshold Demo
Saved, I really like your site!
If your articles are always this heflpul, I’ll be back.
I ralely appreciate free, succinct, reliable data like this.
How neat! Is it rellay this simple? You make it look easy.