blueChannelReuction.py

# from PIL import Image
import numpy as np
import os
import cv2 as cv


tifs = [x for x in os.listdir('./T152_Full') if '.tif' in x]

def threshold_blue(image_file, threshold):
    img = cv.imread('./T152_Full/' + image_file, 1)
    rows, columns, channels = img.shape


    img[np.where((img < [threshold,255,255]).all(axis=2))] = [0,0,0]
    
    # for i in range(0, rows):
    #     for j in range (0, columns):
    #         blue_px_val = img.item(i,j,0)
    #         if blue_px_val <= threshold:
    #             new_img[i,j] = (0,0,0)
    #         # else:
    #             # new_img[i,j] = img[i,j]

    cv.imwrite('./processedBlue/'+ image_file, img)


if __name__ == '__main__':
    for t in tifs:
        print(t)
        threshold_blue(t, 75)










# outputs a list of "pixel # : frequency " values
# the data actually appears for pixel values that's a multiple of 5
def create_histogram(image_file):
    img = cv.imread('./comparedata/'+image_file, 1)

    rows, columns, channels = img.shape

    # blue_px = img[:,:,0]

    hist = cv.calcHist([img],[0],None,[256],[0,256])
    vals = range(0,255)

    # extracts values from a list of lists
    frequencies = [freq for px_val in hist for freq in px_val]
    # print(frequencies)
    
    for i in range(0,len(frequencies)):
        print(i, frequencies[i])