An Imaging Mechanism Based Method For Optical Remotely Sensed Data Simulation And Spatial Scale Calibration

Multi-source remote sensing data fusion has become one of the main trends inremote sensing applications, which many studies have focused on innovation anddevelopment in a variety of fusion methods or techniques. However, the differences inthese multi-source data of the internal system are considered less. These differencesare mainly caused by different sensor design specifications for data, such as differentspatial resolution, different spectral response function, etc. Before the application ofdata fusion technology, we need to evaluate these differences, then the consistencycorrection, in order to improve the fusion quality. Firstly, from the angle of the opticalremote sensing imaging mechanism, the spectral and spatial response of the imagingprocess simulation are carried out simultaneously, and simulation methods areevaluated; these methods are applied to simulate different remote sensing data (ETM+,MODIS, AVHRR3), and evaluate their consistency, and may ultimately based on dataconsistency correction.As an important application example of multi-source remote sensing integration,high-resolution data are often used to the scale effect correction to the low spatialresolution data. We use simulated high and low spatial resolution remote sensing data(no inherent system differences) to implement the scale effect correction. Then, avariety of current correction methods are summarized, analyzed and discussed fortheir respective advantages and shortcomings, and finally some specific methodselection strategy are proposed in this study.According to the contents and findings of this study, we obtained the followingconclusions:(1) In the multi-source remote sensing fusion process, different remote sensingdata consistency evaluation and calibration are required. Simulation techniques basedon optical remote sensing data of the imaging mechanism can effectively solve thisproblem, and provide the basis for data simulation and the design of sensors withdifferent spectral and spatial characteristics of the data;(2) The heterogeneity of regional remote sensing applied research, take fullaccount of the scale effect of the remote sensing data. Integrating Taylor expansionformula and the discrete variance of a pixel will improve scale effect of biascorrection; variogram combination of integral range (A), can effectively choose theoptimal scale of the study area; (3) Different remote sensing scale effect correction methods have specificapplication conditions, in the practical application. It is necessary to consider theadvantages and disadvantages of various methods as well as objective conditions andchoose the best scale effect correction algorithms.