1.輸入是經(jīng)過(guò)Jet處理的偽彩色圖像，輸出是深度圖

def Jet2Gray(false_color_image):
    gray_values = np.arange(256, dtype=np.uint8)
    color_values = map(tuple, cv2.applyColorMap(gray_values, cv2.COLORMAP_JET).reshape(256, 3))
    color_to_gray_map = dict(zip(color_values, gray_values))
    # apply the inverse map to the false color image to reconstruct the grayscale image
    gray_image = np.apply_along_axis(lambda bgr: color_to_gray_map[tuple(bgr)], 2, false_color_image)
    return gray_image

2.有時(shí)候偽彩色圖像經(jīng)過(guò)編碼等會(huì)破壞原來(lái)的映射關(guān)系，直接用上述方法處理行不通，所以需要經(jīng)過(guò)濾波處理，方法如下：

import cv2
    import numpy as np
    # create an inverse from the colormap to gray values
    gray_values = np.arange(256, dtype=np.uint8)
    color_values = map(tuple, cv2.applyColorMap(gray_values, cv2.COLORMAP_JET).reshape(256, 3))
    color_to_gray_map = dict(zip(color_values, gray_values))
    keys = list(color_to_gray_map.keys())
    pass
    pic_path = "test.png"  # 圖片路徑
    img = cv2.imread(pic_path)
    for x in range(img.shape[0]):  # 圖片的高
        print(x)
        for y in range(img.shape[1]):  # 圖片的寬
            px = img[x, y]
            # print(px)  # 這樣就能得到每個(gè)點(diǎn)的bgr值
            d_min = 100000
            best_val = px
            for key in keys:
                d = 0
                for i in range(3):
                    d += abs(float(px[i])-float(key[i]))
                if d < d_min:
                    best_val = np.array(key)
                    # temp = img[x, y]
                    d_min = d
            img[x, y] = best_val