基于稀疏分解的交通图像压缩

Traffic Image Compression Based on Sparse Decomposition

随着道路实时监控系统的广泛应用,需要处理的数据量激增,为了解决传统图像压缩方法随压缩比增大解码图像质量急剧下降的问题,提出一种新的基于稀疏分解的交通图像压缩算法。该算法将稀疏分解引入到交通图像压缩中,先根据交通图像特点,引入背景差法对交通图像进行预处理,然后采用稀疏分解算法对图像进行分解,最后通过分析交通图像稀疏分解后的数据分布规律,并根据低比特率图像压缩要求,提出改进的排序差分编码方案进行编码,以期在低比特率下,获得较好的解码图像质量。仿真试验结果表明,与现有的排序差分编码算法相比,该算法能够更有效地实现交通图像的压缩,相同压缩比下,解码图像有更高的峰值信噪比和主观图像质量。

Along with the extensive application of the real-time monitoring system,the amount of data which requires processing increases sharply.In order to solve the problem that the quality of the decoding picture falls suddenly with a rise of the compression ratio when using the traditional methods of image compression,a kind of new traffic image compression algorithm based on sparse decomposition was proposed,in which sparse decomposition was introduced into traffic image compression.First of all,according to the characteristics of the traffic image,background subtraction was introduced in pre-processing traffic images.Then the sparse decomposition algorithm was adopted in image processing.Afterwards,the distribution of sparse decomposed data of traffic images was analysed.At last,an improvement of the order differential coding schemes was proposed to get better decoded image quality at low bit rate.The experimental result shows that the algorithm can achieve a more efficient traffic image compression compared with the existing order differential coding algorithm.The decoded image＇s peak signal-noise ratio and subjective quality are better at the same compression ratio.