Image Compression Using an Improved Optical Inner Product Processor

We show a prototype of a coding system for image compression using vector quantization with an optical inner product processor, whose accuracy is improved by the applying data transformation method. We also present compression effects for gray-scaled images by this system. The system can compress 256X256X8 bits image data within 1.2 seconds under 0.18 compression ratio with an adequate quality.