Morphology Analysis Of Asteroid Remote Sensing Image Based On Hierarchical Classification Enhancement Algorithm

In recent years,the successful navigation of various space probes has acquired a large number of asteroid remote sensing images.The asteroid's topographical features are no longer mysterious.The asteroids existed in the early days of the sun,and some material information about the early solar system remains.Asteroid research can enhance the understanding of the structure of the solar system.Studing the shape characteristics of asteroids is of great significance to the selection of the navigation,prevention of celestial bodies from colliding with the Earth,development of extraterrestrial resources,and obtainment of extraterrestrial wealth.Therefore,the analysis of the topographical features of asteroids has become an vital scientific issue in current research.Processing of asteroid remote sensing images is an essential step in the analysis of asteroid's topographical features.The high spatial resolution images of near-Earth asteroids provided by space probes give us opportunity to observe the surface of the asteroid carefully.However,due to the tough shooting conditions,the remote sensing images are usually affect by the light intensity and shooting angle,and the quality is relaticely low.The low contrast areas in the asteroid imageseriously restrict the accuracy of morphology recognition.For the recognition of low-contrast morphology in asteroids,we need to enhance the remote sensing image.Existing image enhancement methods such aslinear stretching,logarithmic stretching,histogram equalization,etc.are commonly used in remote sensing.The image enhancement method can achieve good effect on some asteroid remote sensing imagesand the shape features of some asteroids can be clearly reflected.However,these methods have some shortcomings:(1)The traditional image enhancement method mainly enhances the image by contrast stretching,but the stretched pixel value is limited,and the human eyes has no resolution to the closer pixel value;(2)The traditional image enhancement method not only has limited enhancement effect on asteroid remote sensing images with low contrast area,but also loses part of the information of the illumination area,which is not conducive to the identification of asteroid shape features.In consideration of the above problems,relative research are carried out and presented in this thesis as follows:(1)A hierarchical classification enhancement method based on asteroid remote sensing image.Firstly,the idea of hierarchical classification is introduced,that is,dividing hierarchical classificationinto macro level classification and micro level classification,and then conduct hierarchical classification experiments on Eros images at the macro level and micro level respectively.Hierarchical classification experiments verify that the hierarchical classification enhancement method is effective for the enhancement of asteroid remote sensing images with low contrast regions.Experimental results show that this method can not only provide a new perspective for analyzing previously reported morphology objects,but also reveal some new features of the low contrast region.That is,revealing the new topographic features of Eros:a connected region with specific spectral values at the asteroid's contour.Finally,using the existing image enhancement methods and the hierarchical classification enhancement method for comparison experiments,The experimental results show that the hierarchical enhancement classification method has better effects in the low contrast region than the existing image enhancement methods and can make it easier to identify the aeromorphic features of the asteroid.(2)The application of hierarchical classification method on asteroid remote sensing image and the analysis of asteroid's topographical features.First,the hierarchical classification method was used to classify all the remote sensing images of near-Earth asteroids.This method is also available for asteroid images with low-contrast region,and able to reveal the connection region of the specific spectral value at the asteroid's contour as well as the new features of some asteroids with fixed stars.Then,the morphology of the asteroids were analyzed according to the hierarchical classification results combined with the obtained asteroid remote sensing images,and five typical characteristics such as craters,megaliths,hills,gullies and ridges which are common on the asteroid surface were summarized.