Infrared Images Of The Dm642-based Real-time Entropy Encoding And Decoding System

This paper designs a system which can realize real-time compression coding, network transmission, decoding and the initial image enhancement for the 320×256 infrared images.Firstly, the analysis of the features of the infrared images was done. Meanwhile, the comparison of the performance of the conventional images compression ideas was made. The combined algorithm of run-length and adaptive Huffman coding was chosen for the system.Secondly, the hardware based on TMS320DM642 was used as the compression coding platform. The network communication function between the processing platform and PC was designed. The PC machine was used as platform for decoding.Thirdly, the combined algorithm of run-length and adaptive Huffman coding based on TMS320DM642 was designed. The improvement and optimization work was also done to the algorithm. The data-sending program based on NDK and the data-receiving program based on Windows Socket was designed according to the principle of network communication. By this means, DM642 can send the compressed date to PC via the net cable. The decoding program based on Borland C++ Builder was designed to decode the images and enhance the initial images. And the needed images can be saved if necessary.Finally, the whole system was debugged successfully and the amount of image data is reduced, the number of the transferred frames per second was increased. And the goal of real-time coding-decoding of the infrared image was achieved.