The Process Flow Research And Application Of Unifier Ray And Color For The Aerial Remote Sensing Image

The phenomenon of cloud、color distortion、shadow in aerial remote sensing imageresults from camera condition、different weather and terrain limitation. Thesephenomenon is the direct acting factors of image after geometric correction in visualeffect and reading effect, what can sufficiently reducing the effect of task execution andit’s result, seriously overrun the impact of traditional pattern noise on image quality.Onlyunifier the ray and color of image by processing factors affecting one by one, can thelatter task execution of digital image keep on doing and ensure the result of the final “oneimage”. Although the study of image quality is the hot research area and some singlequality problem had been studied out, such as removing image shadow and cloud inimage and so forth, these is no thorough process flow can handle various kinds of factor.This essay propose a process flow of unifier ray and color for remote sensing image,which on the basis of DOM processing, mainly on Handling of quality issue thoroughlyand one by one. The main flow is: firstly, according to the original ground resolution ofimages and the national mapping scale,resample in the appropriate sampling methods;secondly, handle the single image factors after appropriate resample, including cloudremoval, shadow elimination and color distortion reduction; thirdly, adjusting theconsistency of ray and color for all the images after handled the single factors; finally,adjusting the trend and feather amount of seam line to eliminate the line trail on themosaic image. Then, the perfect result of mosaic image has been completed. Fifteenimages, which from the same coverage of survey area, will be used in the essay to be thetest demonstration to get the experiment results by different traditional and now availableimage processing software of remote sensing. The contrast of the result images and theimage quality assessment information demonstrating that the processing flow proposedby author is much more better than traditional and now available image processingsoftware of remote sensing, meanwhile,it covers the shortage of single method and canform the solution for massive data, the final result from which can meeting therequirements on national standards of DOM image.