python中將nii窗寬窗位歸一化，並轉為jpg

import numpy as np
import os  # 遍歷文件夾
import nibabel as nib  # nii格式一般都會用到這個包
import imageio  # 轉換成圖像

center = 500 # 肺部的窗寬窗位
width = 1000
# print(center)
# print(width)


def nii_to_image(filepath):
    filenames = os.listdir(filepath)  # 讀取nii文件夾

    for f in filenames:
        # 開始讀取nii文件
        img_path = os.path.join(filepath, f)

        img = nib.load(img_path)  # 讀取nii
        img_fdata = img.get_fdata()  # api 已完成轉換，讀出來的即為CT值
        fname = f.replace('.nii.gz', '')  # 去掉nii的后綴名
        img_f_path = os.path.join(imgfile, fname)
        # print(img_f_path)
        # 創建nii對應的圖像的文件夾
        if not os.path.exists(img_f_path):
            os.mkdir(img_f_path)  # 新建文件夾

        # 轉換成窗寬窗位
        min = (2 * center - width) / 2.0 + 0.5
        max = (2 * center + width) / 2.0 + 0.5
        dFactor = 255.0 / (max - min)
        # print(dFactor)
        # #
        # #     # 開始轉換為圖像
        (x, y, z) = img.shape
        # print("xyz",x,y,z)
        for i in range(z):  # z是圖像的序列
            silce = img_fdata[:, :, i]  # 選擇哪個方向的切片都可以
            # rawData_win = np.zeros(rawData.shape, dtype='float32')

            silce = silce - min
            silce = np.trunc(silce * dFactor)
            silce[silce < 0.0] = 0
            silce[silce > 255.0] = 255  # 轉換為窗位窗位之后的數據

            imageio.imwrite(os.path.join(img_f_path, '{}.jpg'.format(i)), silce)


if __name__ == '__main__':
    filepath = 'D:\good good study\data analyse\cervical cancer_deeplearning\\bilibili_LEO\preprocessing\T2_test'
    imgfile='D:\good good study\data analyse\cervical cancer_deeplearning\\bilibili_LEO\preprocessing\T2_test_png'
    nii_to_image(filepath)