Application Research Of Correspondence Points Extraction Method Based On Crater’s Edge And Its Use In Chang-e Ⅱ Images

Compared with Chang-e I CCD images, The Chang-e II has higher spatialresolution, that also means that its CCD camera has a narrow viewing angle, and itsimages are more vulnerable with the sensor attitude/position, perspective projectioncenter, panoramic distortion, surface topography and other factors, to cause the imagedistortions such as geometric distortion, scale changes, ground displacement, whichhave adverse effects to the feature extraction, craters recognition, etc.so its quiteessential to Ortho Corrected the Chang-e II images. according to whether or not to useDEM, orthorectification can be divided into DEM based and DEM-independentOrthorectification. DEM based Orthorectification also can be divided into singleorthorectification and full digital photogrammetric system, the later method is widelyused in industrial digital photogrammetry processing because of its fast operation,high treatment efficiency, but it’s not so expansive and not convenient at all. Therefore,we often adopts single-chip correction method when dealing with the single image.The DEM extracted by digital photogrammetric methods from Chang-e onethree linear CCD image, which has500m spatial resolution, is the highest DEMspatial resolution products in present days. but its spatial resolution is far frommeeting the needs of orthorectification of Chang-e II CCD image’s7m spatialresolution. Therefore, through the three-dimensional relative positioning of Chang-e IICCD stereoscopic, image matching, etc. this paper extract a high precision digitalelevation model, And in this process, this paper come up with a new correspondencepoints extraction method: correspondence points extraction method based on Entity’sedge corresponding points extraction methods based on a Laplace operator andgray point, and the extraction accuracy and extraction efficiency was verified byimage experiment; and this paper selected rational polynomial orthorectificationmethod to Ortho Correct the Chang-e II CCD images to eliminate the adverse effectsof sensor tilt, satellite attitude changes, panoramic distortion, perspective projectionand others, to obtain a spatial resolution orthorectification image which can reflect thereal lunar surface topography information; Also, this paper described the geometriccharacteristics of craters based on the six-figure crater quantitative classificationmethod. The main contents of this paper includes: (1) Chang-e I CCD data preprocessing. Including image selection, geocoding,image de-noising.(2) Correspondence points extraction method based on Entity’s edgecorresponding points extraction methods based on a Laplace operator and gray point,and the extraction accuracy and extraction efficiency was verified by imageexperiment.(3) DEM extraction. Including Chang-e I laser altimeter data processing, thelunar surface control points selection, the control points inspection and verification,external orientation elements calculation, image matching, DEM generation andchecking.(4) Orthorectified. Including the exterior orientation elements calculation, addingterrain data, generate ground control points, orthophoto generation,etc.(5) Orthophoto applications. partial lunar image six-figure crater quantitativeclassification applications.On the basis of the above research, innovation and research results of this paperincludes:(1) A method of Chang-e II I CCD image matching and corresponding imagepoints extracting methods based on a Laplace operator and gray point is created.Andthe the accuracy and efficiency of the methods in image matching and correspondingimage points extracting is verified by compared with the methods based on features.(2) This paper selected rational polynomial method to extract high spatialresolution DEM.(3) By the use of the DEM, this paper selected rational polynomial method toOrtho Correct the Chang-e II CCD images, and extract high spatial resolution DOM;(4) This paper described the geometric characteristics of craters based on thesix-figure crater quantitative classification method. and got a high accuracy and effect.