Research On Methods Of Stereo Image Compression

As one of the ways for obtaining digital terrain models (DTM), stereo photography is the basis of digital photogrammetry. Nowadays, satellite stereo photography have been applied to earth observation by lots of countries, such as SPOT of France, ALOS of Japanese, IKONOS and QuickBird of USA and CBERS, Chang'E1, Chang'E2 of China. All of them have the ability of stereo photography. Besides mapping and remote sensing, stereo and multiview imaging system have also been used in entertainment area. With the improvement of living standards of people and the development of information technology, the mono-view video can not satisfy the requirement of people any more. Stereo imaging system is applied to entertainment area such as TV and movie. Multiview video and 3D-movies will be a trend of the future of the movies and TV.Stereo or multiview imaging system produces several times more data than mono-view system under same resolution. However, there exists high correlation between each two views. We can lower the bit-rate of pixels by exploiting efficiently the correlation existed in these views to alleviate the pressures of the storage system and transmission system.However, the stereo and multiview image compression technology is still on the primary phase, far from being mature as the common image compression technology. There is still some distance away from the actual application for multiview image compression technology. In this paper, we focus on the compression of stereo and other multiview image by incorporating wavelet transform with adaptive technology and predictive coding theory. The main work includes the following aspects.(1)Considering the good performance and existing problem of vector lifting schemes (VLS), an improved decomposition method is introduced for stereo image compression. The VLS incorporate disparity compensation with the decomposition of target images, in other words, the information of reference image is directly applied in the decomposition of the target image. However, due to the shift-variance property of discrete wavelet transform (DWT), the subbands of reference image can not always provide a good prediction for the coefficients of subbands of target image. As a result, the decorrelation performance of the decomposition of VLS will be greatly affected. In this paper, the relationship of distinct phases of DWT is exploited to eliminate or alleviate the effect of shift-variance to enhance the compensation capability of reference image. In addition, occlusion also incurs a degradation of compensation. So an adaptive method is proposed to solve this problem. Experiments show that the improved technologies lead to a significant coding gain.(2)The effect on the dense match of stereo image derived from image compression is studied by experiments. Stereo image pairs are mainly used to generate DTM, one of the main procedures is dense match. Ability to maintain the accuracy of matching results is an important measure of stereo image compression performance. In this paper, several representative experiments indicate that disparity compensation-based compression method maintains better matching results than independent compression method.(3)After a comprehensive analysis of the existing lossless image compression algorithm, stereo image lossless compression algorithms are proposed in this paper. At present, lossless image compression algorithms consist of predictive coding methods and reversible wavelet transform-based lossless compression methods. Considering the intra-correlation and inter-correlation of stereo image, a lossless predictive coding method is proposed for stereo images. Experiments show that the proposed lossless predictive coding method acquires a good performance by exploiting intra-correlation and inter-correlation jointly. In addition, considering that predictive coding method can not provide a resolution progressive code-stream, another lossless compression method is proposed by incorporating predicting with lifting. This method provides a progressive code-stream for stereo image at the cost of a little loss of performance.(4)In addition to stereo image pairs, three-line-scanner image is also widely used in mapping and remote sensing. In this paper, VLS is extended to the compression of three-line-scanner image. Experiments show that VLS-based compression method is better than spatial disparity compensation and wavelet-based compression method at three-line-scanner image. Furthermore, an adaptive lifting scheme is also proposed for multiview image compression.