Lossy Compression for Raw Satellite Images Using Intra Frame and Inter Frame Coding

Abstract

Abstract:The huge size of the satellite images data leads to big problems in storage capacity and power exhaustion for the transmission, therefore great efforts have been made to find low-complexity compression algorithms. To overcome these issues in this field we have proposed satellite image compression method based on inter-intra frame coding, to compress sequence of satellite images captured at different times, form different sensors, or /and different view points for the same area. Inter-frame coding principle based on removing temporal redundancy exists between two successive satellite images: a historical captured image (reference image) and a new captured image (sensed image) for the same scene. Since sequential satellite images contain almost same materials with only minor movements according to the time interval between them (several days, several months). The resulting difference's image after subtraction has much lower entropy and can be compressed with much less cost than the original sensed satellite image. This is followed by removing still existing redundancy of pixels within resulting difference's image using principle intra-frame coding, and this will be helpful in order to satisfy high compression performance while keeping the quality of the compressed satellite image as much as possible.