I have trained the UNET model and got the PTH file, but how can I process the image to get the segmented image?