Architecture of a Digital Pixel Sensor Array with Tile-Based Vector Quantization Image Compression Algorithm

In this paper, an improved tile-based vector quantization (VQ) algorithm used for image compression application integrated along with a digital pixel sensor (DPS) array is proposed. The proposed technique is based on a tile-based compression of the original image, enabling an easy implementation of vector quantization. Furthermore, tile-based design reduces the accumulation of error inherent in DPCM type of processing. Simulation results show that the compression rate, which is measured by bit-per-pixel (BPP), can be as low as 0.28BPP, with PSNR result around 20dB.