Infrared Image Real-time Compression Based On DSP

In the era of information technology, there is a critical demand on multimedia information such as image ,audio,vedio and so on .digital image is widely used in the fields of family amusement, remote surveillance, network education and public security. But the contradiction is more and more conspicuous between large number of image data and the limited capacity of band width. It is a necessary and hard task for us to research high efficiency arithmetic for image compression. International Standard Organization (ISO) have promulgated several international standards in accordance with various applications, in which the lastest and most efficiency research achievements are adopted, the standards have recommended standard arithmetic and data types. This can direct researcher how to improve and implement the arithmetics. It's the main purpose of this paper to search efficiency arithmetic for infrared-image compression; meanwhile, we aim to implement the arithmetic on TI DSP. Firstly, the common method and hardware device used in image compression were introduced, and several international standards were summarized in this thesis. Secondly, the hardware design was described detailedly, and real-time request of image compression were thought over in the design process. Many detail about hardware, such as TMS320C6211 overview, DSPconnect with peripheral equipment, the setting of DSP core register and CCS2.0 software were involved. Thirdly, forward a control-logic design plan of real-time and multiresolution gathering system which is based on FPGA. Finally, a material arithmetic and the implement were described. Since this arithmetic was based on Discrete Cosine Transfer (DCT) and Huffman coding, fast-DCT, Huffman coding and three step search (TSS) were highlighted in this section. we realized that DWT put up high efficiency in image compression. Analyse the process of DWT. Compare two different real-time image compression results of zerotree coding arithmetic based on DWT and Huffman coding arithmetic based on DCT .