The Improvement Of Micro-satellite's Low-light Remote Sensing Image Quality

The collection of low-light remote sensing in China is too hard to realize.It's practical application is an area of blank.But the obtain of low-light remote sensing is very important to our national defense and military,and it also important to the information commerce.The ability of observing the target anytime we wanted is urgent to be equipped.Nowadays,the improvement of micro-satellite has give us one way to get the low-light remote sensing images.With the low-light sensor,the new satellite platform can staring the earth target longer time than traditional satellite.The paper takes domestic and foreign correlation domain research results as foundation.Make research about this question,working about the possible ways to get it work,and make modes to future study.And also do works about the improvement of the low-light remote sensing collect by this way.New micro-satellite platform can achieve much more longer integration time than the old platform.During the study,the Signal-to-noise Ratio was modeled in math,provides the function relation between the optical relative aperture,integral time and Signal-to-noise Ratio.The satellite platform staring on the earth target can't avoid to vibrate during the integration time.The quality of remote sensing will be declined.So the improvement ofimage quality is necessary.During this part,the basic of evaluation system about low-light remote sensing is made.The platform vibrate is random,so the image recover belongs to the blind deconvolution.By the blurring kernel evaluation,the more accuracy blurring kernel we get,the result of recover will be much better.At the same time,the remove of the image noise is another important part of image quality improvement.Finally,in the laboratory,collect target image,prove the measure we used in image quality improvement is useful.The Signal-to-Noise Ratio and Modulation Transfer Function both have large improve.The result show,the staring model low-light remote sensing has big development space.The image process can make a good improvement of image quality.