question about preprocessing_img

[charlesyou999648]

Hi,
I am pretty interested in your work!
During my implementation, I have a question about preprocess_img in seg_liver.py

def preprocess_img(image, x_bb, y_bb, ids=None):
    in_ = np.array(images)
    in_ = in_.transpose((0,2,3,1))
    in_ = np.subtract(in_, np.array((104.00699, 116.66877, 122.67892), dtype=np.float32))

I applied the following procedures: 1. clip and normalize the image; 2. time 255; 3. subtract the mean for three channels of the images. Then I validated the image but found the image shown below which is not like the usual image as i see before. I am wondering whether my processing is true, could you please leave me some comments? Thanks!