In automotive industry, camera lenses are selected to have a large field of view in order to see both the sides and the far distance in the scene. This results in capturing a warped image as shown. Dewarping is needed to flatten the image before running additional image processing algorithms.
Scope of the project includes comparing different dewarping algorithms with Matlab/Python and analyzing which of them can provide specified dewarping level at the minimum memory consumption. Second target is to implement the algorithm on FPGA with 1 to 9 line buffers and not a frame buffer. Optional target is to develop an even better algorithm.