The output of these sensor layers was then processed in a pretrained convolutional neural network, to generate the final, panchromatic image which could then be used for a wide range of purposes.