节点文献
基于惯性参考基准的像移测量与图像复原
Image Motion Measurement and Image Restoration Based on the Inertial Reference
【摘要】 卫星上存在高、中、低频率的微振动源,引起光学载荷光轴抖动,造成遥感图像成像品质下降。文章提出了一种基于惯性参考基准的像移测量与图像复原方法,介绍了该方法的原理和涉及的关键部件。为验证该方法的可行性和性能,研制了一套地面试验验证系统,包括验证相机(含惯性参考基准单元、哈特曼传感器)、积分球、靶标、平行光管、隔振平台、激振器、六自由度支撑台,以及相应的图像复原软件等。测量了试验环境的背景噪声,开展了像移测量精度试验,在此基础上完成了多种工况下的图像恢复试验。结果表明,基于惯性参考基准的像移测量误差均方根值不超过0.12像元,基于测得的像移数据可显著提升像质,图像的调制传递函数一般可提升至原图的1.38~1.52倍。该方法获得的像移数据还可以反馈给稳像镜,实时补偿星上微振动,有助于直接获取高品质图像。
【Abstract】 There are lots of high,medium and low frequency micro vibration sources on the satellite,which cause optical axis jitter of the optical load,resulting in the degradation of remote sensing image quality.Aimed to that question,a method of image motion detection and image restoration was proposed based on an inertial reference,and the principle and the key components of the method was introduced.To verify the performance of the method,a test system was developed,including a space camera,an inertial reference unit,a Hartmann sensor,an integrating sphere,a simulated image target,a parallel light pope,a vibration isolation platform with the vibration generator,a six degrees of freedom platform,and the software of image restoration.The background noise of the test environment was measured,and the image motion measurement accuracy test was carried out.Based on that work,the verification tests of image restoration under various working conditions were completed.The test results show that the error of image motion detection is less than 0.12 pixels(Root Mean Square,RMS).Using the image motion data to improve the image quality,MTF(Modulation Transfer Function) of the restored image was improved to 1.38-1.52 times of the original image MTF.The image motion data can be used as a feedback to the fast steering mirror to compensate the satellite jitter in real time,which helps to get high quality images directly.
【Key words】 inertial reference; micro vibration; image motion; image restoration; remote sensing satellite;
- 【文献出处】 航天返回与遥感 ,Spacecraft Recovery & Remote Sensing , 编辑部邮箱 ,2021年01期
- 【分类号】V443.5;TP751
- 【被引频次】1
- 【下载频次】112