摘要
根据SAR原始数据经沃尔什变换(WHT)后的系数特性,提出了一种在沃尔什域按各子带能量分布作变比特块自适应量化的算法-沃尔什域块自适应量化(WHT-BAQ),并将本方法与时域BAQ算法和FFT-BAQ算法作了比较。实验证明,在相同比特率下,WHT-BAQ算法的数据域和图像域信噪比均比时域BAQ算法高,与FFT—BAQ算法性能相当。WHT-BAQ算法的运算量略大于时域BAQ算法,但大大小于FFT- BAQ算法的运算量。加之沃尔什变换具有不作乘法、有快速算法的特性,所以将它用于SAR原始数据压缩有一定的优越性。
Synthetic aperture radar(SAR)is an active microwave imaging sensor.It utilizes pulse compression technol- ogy to obtain high range resolution and aperture synthesis technology to achieve high azimuth resolution.SAR has a capa- bility of obtaining high resolution images that no ordinary radar can be comparable with,therefore,it is one of the signifi- cant directions for the modem radar development.The raw SAR data is difficult to be transmitted and stored due to its large size,so it is necessary to use data compression or reduction techniques.In this paper,based on the characteristics of SAR raw data coefficients in Walsh transform domain,an algorithm of variable bit rate allocation according to energey dis- tribution of every subband in Walsh transform domain-Block Adaptive Quantization in Walsh Transform domain(WHT- BAQ)is put forward.The comparison among WHT-BAQ algorithm,time domain BAQ algorithm and FFT-BAQ algorithm is conducted.By using two real SAR raw data,compression and decompression with the three algorithms are made respec- tively.The quality parameters are achieved,and the complexity of the three algorithms for raw SAR data compression are studied.The images correspond to the three algorithms are gained.The experiments manifest,due to bit allocation is introduced according to energy of different subband in Walsh transform domain,therefore,with same bit rate,SQNR and SDNR of WHT-BAQ algorithm surpass that of time domain BAQ algorithm.The increase of I.15-I.48dB is gained for SQNR,and the increase of 1.54-1.73dB is gained for SDNR.The performance of WHT-BAQ algorithm is comparable with FFT-BAQ algorithm.In different bit rate,the variance of resolution.PSLR and ISLR is small for the three algorithms. The above mentioned manifests,data compression technique(in the case of bit rate more than 2 bits/sample)hardly in- fluences resolution of SAR image hardly,and SAR image is not sensitive to bit rate of raw SAR data.The operation of WHT-BAQ algorithm is a little of exceed BAQ algorithm,but is much less than
出处
《遥感学报》
EI
CSCD
北大核心
2008年第3期392-398,共7页
NATIONAL REMOTE SENSING BULLETIN
基金
航空科学基金(编号：05D52027)
关键词
数据压缩
沃尔什变换
块自适应量化
信噪比
比特率
data Compression
Walsh transform
block adaptive quantization
signal to noise ratio
bit rate
