摘要
为了验证星上遥感图像实时非均匀性校正的必要性,对TDICCD像元非均匀性对遥感图像压缩的影响进行了深入研究,并利用FPGA设计了一套实时非均匀性校正系统。首先分析TDICCD像元非均匀性产生原因,以及非均匀性对后续图像压缩影响的原理,并根据星上实际应用的条件,提出了一种保留增益的单点非均匀性校正算法,该方法即保留了2点校正算法适应性强的优点,又克服了星上2个标准源难以实现的缺点。最后进行实验验证,结果表明非均匀性优于1%,并将校正前后的图像分别进行了图像压缩,对压缩效果进行了比对,无损压缩时,像元非均匀性校正后,压缩比提高1.5倍左右;有损压缩时,PSNR提高4 d B以上,表明了星上实时图像压缩的必要性,以及本文系统和算法的可行性。
In order to verify the need for real-time non-uniformity correction of satellite remote sensing image , the effect of TDICCD pixel non-uniformity on image compression was deeply studied , and the real-time non-uniformity correction system based on FPGA was designed.Firstly, the cause of TDICCD pixel non-uniformity and the princi-ple of the effect of TDICCD pixel non-uniformity on image compression were analyzed .And according to the practi-cal application conditions on satellite , a remain-gain single-point non-uniformity algorithm was proposed .The algo-rithm that retains the advantages to highly adaptive, but also to overcome the shortcomings of two standard sources is difficult to achieve on satellite.Finally, the experiment results show that the non-uniformity of better than 1%, and before and after correction images were compressed , the compression effect was contrasted , when lossless com-pression, the compression ratio increased about 1.5 times after pixel non-uniformity correction, when loss compres-sion, PSNR raise more than 4 dB.It indicates the necessity for real-time image compression on satellite and the feasibility of this system and algorithm .
出处
《电子测量与仪器学报》
CSCD
北大核心
2015年第3期375-382,共8页
Journal of Electronic Measurement and Instrumentation
基金
国家自然科学基金(61405191)资助项目
关键词
信息处理技术
非均匀性校正
合成压缩
推扫式遥感相机
多通道TDICCD
information processing technology
non-uniformity correction
composing and compression
push-broo mremote-sensing camera
multi-channel TDICCD