Research On Infrared Image Compression Technology Based On Wavelet

With the infrared image and its obtaining technology development, information of images also shows growth multiples. For their huge volume of data, static infrared images and infrared image sequences are usually difficult to be used in practice, because the large data stream seriously depletes the limited bandwidth resources. A high-speed and real-time transmission for UAV (unmanned aerial vehicle) infrared images has been a"bottleneck"of the UAV's application, so that a rapid and efficient transfers for infrared images has attached great importance to the infrared field at home and abroad all the while. Today's digital battlefield makes the development of infrared image compression is of great significance.This paper first understands the existing standards and algorithm of image compression. Currently, most infrared image compression systems adopt the scheme of existing mature image and video coding standards. The research on image compression arithmetic is a hotspot, there are many mature algorithms have been applied to practice, however, each of them has its own applicability and limitations. Infrared images have inherent particular characteristics, so a common image compression method for infrared image may not be the most appropriate choice. Therefore, we must combine the characteristics of infrared image compression algorithms for the study, and set higher requirements for it.Secondly, base on the theory of wavelet transform, this paper introduces several commonly used wavelet functions and how to select the wavelets. Then it analyses the key technologies of image and video compression while combining the existing coding standards, and brings in a more effective algorithm aim at the characteristics of UVA infrared image sequence. For static image, the paper studies two widely used algorithms: Embedded Zero-tree Wavelet Encoding (EZW) and Set Partitioning in Hierarchical Trees (SPIHT), and realizes an integrated improvement for SPIHT aim at the infrared image and wavelet filter characteristics on the base of the study. For image sequence, this paper proposes a scheme of UAV image sequence compression and realizes it on base of the static image compression coding algorithm before. Finally, we simulate the improved algorithm and analysis it. The result shows that the improved SPIHT algorithm gets a PSNR 0.3-1.0 dB better than the original SPIHT under the same compression ratio condition. The image sequence compression scheme adopts adaptive motion estimation methods, so that the adaptive frames selecting can maximize the use of time-relation, and comparing with the H.264 standard, the average PSNR per frame of the restored image sequence improves 0.1dB. This method can advance the capacity of error code resistance, reduce the computing amount of motion estimation, and decrease the computational complexity and saving rate to meet the requirements of real-time and high-speed.