An error detection and recovery technique for images compressed with the CCSDS compression algorithm

Variable-Length Codes (VLC) are highly sensitive to transmission errors. Even a single bit error may de-synchronize the decompression that can lead to severe image quality degradation. In this paper, a novel error detection and recovery technique for the reception of compressed image sequences containing several bit errors is proposed. The image data compression algorithm is based on the Consultative Committee for Space Data Systems (CCSDS) recommendation for space systems. In the event of de-synchronization due to bit errors, our proposed technique will exploit the syntax and semantics of the compression algorithm by first identifying the elements in a compressed sequence that can potentially cause the de-synchronization. These elements are then processed in a methodological manner in an attempt to re-synchronize and recover the remaining uncorrupted sequence. Simulation results indicate a significant reduction in the Mean Squared Error (MSE) of the reconstructed image when our proposed technique is invoked.