Generalized error diffusion method for halftoning

Halftoning is a process to produce binary image having gray shade information. There are several halftoning methods, which are implemented in different ways and we have to write separate programs for the different methods. It has been found that the halftoning methods are based on a common principle of distributing error in the neighboring unprocessed pixels. So, it can be implemented in a single kernel based method, where the kernel specifies the amount of errors to be distributed to neighboring pixels. This method is simple to implement and fast to execute while maintaining the same performance as the original methods. Also, it allows us to find better kernels which give better halftone images.