Design And Realization Of Remote Sensing Image Data Processing Platform

With the develop ment o f modern phys ics, space technolo gy, e lectronictechno logy, co mputer techno logy, information sc ience and Enviro nmenta l Scie nce,remote sens ing techno logy has become an advanced and practica l means ofcomprehe ns ive detection, it has been widely used in vario us fie lds, such as agriculture,forestry, geology, ocean, hydro logy, meteorolo gy, geography, enviro nmenta lmonitoring, Earth resource exploration and military reconna issance. Therefore,analytica l processing re mote sens ing images that are better and more accurate, canprovide better decis ion support in response to the s ituatio n of c limate variability,natura l disasters, and ecologica l destruction and so on.This paper relies on the project undertaken by CEODE, builds the platform ofremote sens ing image data processing, has a research on a series of processingmethods for hyperspectral re mote sens ing data and disc usses the key technolo gies,inc lud ing as the fo llows:Firstly, the paper introduces several algorithms for remote sens ing radio metricca libration progra m, through the comparison o f several ca libratio n method, designsthe most appropriate solutions, and allows the users to generate the calibrationprogra m the ir own needs by d ifferent s ituations.Secondly, the radiation properties of the re mote sens ing ima ges are eva luatedfro m the two parts of the calc ulations of the instanta neous fie ld of view angle andgenera lized no ise. In the calculation o f instanta neous fie ld of view angle, accord ing tothe monitoring and forecasting o f s mall satellite o f environment and disaster, theauthor designed a relatively simp le calculation method which can be better to get theresults o f the insta ntaneous fie ld of view angle. The part of genera lized noiseca lculation uses the currently most commonly used computing model to ana lyze，andgets the genera lized no ise results by a series of mathe matical derivation.Then, by the spectral ima ge noise eva luatio n and texture ana lys is to evaluateima ge qua lity. In the noise eva luatio n, through the design of s igna l to noise ratioca lculations and Imp ulse noise statist ics, there is a reasonable eva luation of the noise;in the part of the image texture ana lys is, designing the calc ulation of the four textureparameters to measure the smoothness of the ima ge.Lastly, the paper inspects the authenticity o f the invers ion products. Theinvers ion of the spectral surface albedo is designed, and thus inspecting the authentic ity of the invers ion products by matching processing to provide a guaranteefor data credib ility. Fina lly there is a summary o f the main work of this artic le.