Abstract: A pixel array of an image sensor includes a plurality of color filter array (CFA) cells. Each of the plurality of CFA cells includes first, second, third, and fourth CFA blocks. Each of the first and second CFA blocks includes a plurality of pixels to sense light of a first color. Most pixels of the third CFA block are configured to sense the second color and the third CFA block includes at least one first pixel to sense the first light. Most pixels of the fourth CFA block are configured to sense a third color of light and the fourth CFA block includes at least one second green pixel to sense the first light.

Abstract: Methods and apparatuses for correcting bad image pixels are described. The described sensor-independent image processing techniques leverage one or more dynamic dictionaries of learned filters for bad pixel correction (e.g., where a camera leverages such dictionaries to efficiently identify filters to accurately adjust and correct bad pixel values). For example, a dictionary may store filters that are learned offline (via a self-supervised learning algorithm implemented at a server using known images and ground truth bad pixel correction values). To select a filter for a bad pixel correction operation, a camera may encode an image patch surrounding a bad pixel (into an encoded patch descriptor) and search the dictionary for a matching patch descriptor key. The camera may then apply the filter (value) corresponding to the searched patch descriptor (key) of the dictionary to the image patch to correct the bad pixel and generate a corrected output image.