On-line filter generation for binary image processing using FPGA

Image processing technology is widely used in various environments. To realizing image processing technology, design of image filters is important. To apply image processing to various applications, application dependent designs of the image filters are required. In changing or unpredictable environments, the image filters have to be modified correspondingly. Therefore, an adjustable filter system is proposed. The system finds the best image filter using a search algorithm. The image filter consists of a set of 3×3 binary matrices. Binary images are considered. The desired filter set is generated by genetic algorithms (GAs), where a set of 3×3 binary matrices is regarded as a chromosome. Simulations are carried out to show the effectiveness of the proposed system. The system is implemented on hardware (field programmable gate array (FPGA)), to keep up with the changing environments. Emulations show that execution time on the FPGA is 48.6 times shorter than that on software