batch pre-process.py

import imageio
import scipy.ndimage as ndi
import matplotlib.pyplot as plt
import numpy as np
from PIL import Image
import cv2 as cv
import os
i = 0
inputfiles = os.listdir('D:\wp\Dataset2023.1.2/new dataset\patient21&black tumor group/')
for inputfile in inputfiles:
    im = imageio.v2.imread('D:\wp\Dataset2023.1.2/new dataset\patient21&black tumor group/'+inputfile)
    hist = ndi.histogram(im,min=0,max=255,bins=256)
    cdf = hist.cumsum()/hist.sum()
    print(hist.shape)
    print(cdf.shape)
    #plt.plot(hist)
    #9plt.plot(cdf)
    #plt.show()
    im_equalized = cdf[im]*255
    hist2 = ndi.histogram(im_equalized,min=0,max=255,bins=256)
    #plt.plot(hist2)
    fig, axes = plt.subplots(2,2)
    im_tissue = np.where(im>90,im,0)
    im_tissue2 = np.where((im_equalized>80)&(im>90),im_equalized,0)
    m = np.where(im>150,1,0)
    im_tissue3 = ndi.binary_erosion(m,iterations=5)
#axes[0,0].imshow(im)
#axes[0,1].imshow(im_equalized)
    #axes[1,0].imshow(im_tissue2)
#axes[1,1].imshow(im_tissue2)
    img = im_tissue2
    cv.imwrite('D:\wp\Dataset2023.1.2/new dataset\patient21&black tumor group/'+"{:0>3}".format(str(i))+'.png',im_tissue2)
    i +=1
    #plt.show()