Thank you very much for your answer. I would like to ask, when both my input and output images are 512 (not 96 or 112), is those points still applicable? (My idea is to scale up proportionally.) My idea is to scale up proportionally, but I'm not sure if there are any issues with doing so?

src = np.array([ [30.2946, 51.6963], [65.5318, 51.5014], [48.0252, 71.7366], [33.5493, 92.3655], [62.7299, 92.2041] ], dtype=np.float32 )