Changing the shape of an image to match dimensions of input shape of model

Question

I trained a model that expects an input shape of (None, 9, 100, 100, 3). When I run the following code:

                    start_frame = 1