| The dense atmosphere on the earth’s surface is one of the important factors hindering the recognition of ground objects,when remote sensing satellites observe ground objects.The transmission of solar radiation signals between the ground,the sun and the sensor will be attenuated by molecules such as aerosols and absorptive gases in the atmosphere,resulting in changes in quality parameters such as image contrast and sharpness.Therefore,atmospheric correction is a key link in the data processing of optical remote sensing sensors.The quality of the atmospheric correction effect directly affects the accuracy of target recognition,sensor performance evaluation,and the inversion accuracy of atmospheric parameters.Based on the 6S(Second Simulation of the Satellite Signal in the Solar Spectrum)radiation transmission model and MODIS auxiliary data,this paper improved the atmospheric correction algorithm to remove the influence of atmospheric absorption and scattering on the image.Since the intuitive manifestation of atmospheric influence is the changes of image quality,in order to verify the influence of atmospheric correction on image quality,the atmospheric radiation transmission was quantitatively linked with the evaluation of on-orbit satellites image quality by analyzing the image signal-to noise ratio,sharpness and modulation transfer function before and after correction.It was concluded that the improved correction algorithm can effectively improve the image quality.The main contents of this paper are as follows:1.Based on the 6S atmospheric radiation transmission model and the MODIS aerosol optical thickness data,atmospheric correction was performed on the atmospheric top radiance images of the four bands of the Gaofen-1 satellite wide field of view camera,and the spectral response curves of different objects before and after the correction were compared.Comparing the correction results of this paper with the FLASSH(Fast Line-of-sight Atmospheric Analysis of Hypercubes)correction results,the results showed that the atmospheric correction algorithm in this paper is good,and it provides a feasible way for the high efficiency and precision of atmospheric correction.2.Aiming at the problem that the signal-to-noise ratio of remote sensing images is easily affected by the boundary among complex ground objects,a signal-to-noise ratio evaluation algorithm based on the principle of threshold segmentation was proposed.The SNR of the GF-1WFV data before and after correction and the synthesized data in band 1,3,and 4 of MODIS1 B were evaluated by using this algorithm,and the results showed that the signal-to-noise ratio of each band before and after correction increases with the radiance.The atmosphere has a greater influence on the short-wave band and less on the long-wave band.The signal-to-noise ratio of MODIS data also increases with the radiance.The verification showed that the method is stable and reliable,and can provide important reference information for the application of remote sensing images.3.Based on the improved Tenengrad gradient algorithm,intermediate frequency-discrete cosine transform method and wavelet transform gray-gradient co-occurrence matrix,these three sharpness evaluation methods counted the sharpness value of the image before and after correction.The analysis and calculation results showed that the image is clearer after atmospheric correction.The sharpness value after correction was higher than the before correction,so the edge texture of the image after atmospheric correction is clearer than before correction.On this basis,the comprehensive image quality evaluation index of the optical satellite sensor-modulation transfer function MTF in-orbit was tested.The knife edge method was used to detect the lakeshore line extracted from the Gaofen-1 image,and the MTF value at the Nyquist frequency before and after correction was compared.The blurring effect of the atmosphere on remote sensing images leads to a decrease in MTF value.The systematic evaluation of remote sensing image quality in this paper provides new ideas for subsequent research. |
PDF Full Text Request