We can make hybrid images from existing images by extracting the low or high pass component from an image and generating the missing component. We are effectively solving an inverse problem.
