基于提升小波变换的雷达视频数据实时压缩算法

雷达视频数据压缩具有大数据量、需要高保真度和实时处理的特点.而现有的一些高保真数据压缩算法都比较复杂,不能满足实时处理的要求.已有的硬件实现算法虽然可以实时处理,但其保真度差,同时难以调整压缩比.通过对实际雷达视频回波数据的分析,表明该数据具有相关性强、相关长度短的特点.在此基础之上,本文提出了一种基于提升格式5-3小波变换和简单的Golomb-Rice编码方法的压缩算法,该算法具有低复杂、高保真的特点,并对算法的运算量进行了详细地分析.实测实验表明,通过软件方式实现该算法即可满足雷达视频数据实时压缩的要求.最后,该算法已被成功应用到雷达海情记录系统中.

A Real-Time Radar Video Data Compression Algorithm Based on Lifting Wavelet Transform

High-quality and real-time processing are two important requirements for the compression of large radar video data set.However,the existed high-quality compression algorithms are too complex and time-consuming for the real-time processing.Some algorithms implemented by hardware can meet the real-time processing requirement,but their compression quality is low and the compression ratio can't be adjusted.In this paper,the characteristic that radar video data has strong correlation and short correlation length is obtained by analyzing actual radar data.Then a low-complex and high-quality compression algorithm is proposed based on lifting wavelet transform with length 5-3 and Golomb-Rice coder.The computational complexity of the algorithm is analyzed in detail.Experiment shows that the software implementation of the algorithm can meet the requirements of real-time radar video data compression.Finally,the algorithm has been successfully utilized in a record system of coastal surveillance radar.