图像的随机翻转和RGB抖动/轻微变化？

Question

I want to implement a small program which will randomly flip and introduce RGB jitter/slight value change.

And if possible to limit the jitter/slight value change to 2 of the 3 layers in the color image.

import cv2
import random

probofflip=0.5
probofRGBjit= 0.6

img=cv2.imread('path/to/img.png',1)
if (random.uniform(0,1)>1-probofflip):
    img= cv2.flip(img,1)
if if (random.uniform(0,1)>1-probofRGBjit):
    #function to jitter the RGB layers 
#do something with resultant image.

Answer 1

If you are using cv2, then numpy is quite helpful for this kind of operation. By jitter, do you mean shifting some pixels around? This example only deals with slight value change.

import cv2
import numpy as np
from pylab import *

img = cv2.imread( r'C:/Users/Public/Pictures/Sample Pictures/Penguins.jpg' )
img = cv2.cvtColor(img, cv2.cv.CV_BGR2RGB)  # cv2 defaul color code is BGR
h,w,c = img.shape # (768, 1024, 3)

noise = np.random.randint(0,50,(h, w)) # design jitter/noise here
zitter = np.zeros_like(img)
zitter[:,:,1] = noise  

noise_added = cv2.add(img, zitter)
combined = np.vstack((img[:h/2,:,:], noise_added[h/2:,:,:]))

imshow(combined, interpolation='none')

If you want to shift each color channel by some pixels, then you can use np.roll. ex:

# shift each channel by 10 pixels
R = img[:,:,0]
G = img[:,:,1]
B = img[:,:,2]
RGBshifted = np.dstack( (
    np.roll(R, 10, axis=0), 
    np.roll(G, 10, axis=1), 
    np.roll(B, -10, axis=0)
    ))
imshow(RGBshifted)