Research On Star Image Preprocessing And Star Point Extraction Technology Based On Star Sensor

At present,due to the high precision,high autonomy,and high reliability of inertial/celestial navigation,it is often used in high-precision navigation systems.In celestial navigation,star sensors are widely used as attitude sensitive devices that can output attitude information in real time.celestial navigation matches the star database with the relative position of the star point in the star map,and then solves to obtain the attitude information.Therefore,the accuracy of the star point coordinate extraction in the star map will directly affect the accuracy of navigation.The accuracy of star point coordinate extraction is determined by many factors,the most important of which is the noise of the star map and the accuracy of the star point extraction algorithm.Therefore,it is of great significance to study the denoising technology of star map and the extraction technology of star point.This article studies the preprocessing and star point extraction technology of star maps taken by star sensors.The main research contents are as follows:First,based on the structure and working principle of the star sensor,this paper analyzes the sources of noise in the star map,uses pure software simulation and semi-physical simulation platform simulation to obtain the noisy star map,and analyzes the star point diffusion model.Through the simulation of various traditional denoising methods,the advantages and disadvantages of the star denoising method are analyzed and summarized.A denoising method combining mathematical morphology and wavelet analysis is proposed.After simulation verification,the algorithm in this paper can retain the shape and gray value of the star point well.Secondly,in view of the motion blur phenomenon of the star map generated during high dynamic operation,an improved R-L algorithm is proposed to restore the star map.When the carrier is running at high speed,since the image shooting is not completed in an instant,the displacement of the carrier during the exposure time will cause the energy of the star point to diffuse,so that the shape and gray value of the star point will change.It is difficult to obtain the coordinate information of the star point.Then the effect of navigation will be affected.According to the current running speed of the carrier and the exposure time of the star sensor,the blur length can be calculated.According to the analysis of the frequency spectrum of the motion blur star map,the Fourier transform can be performed on the star map twice to obtain a more accurate blur angle.After the degradation function of the blurred image is obtained,the R-L algorithm is used for restoration to obtain the star map after the shape and gray valuerestoration,which can effectively improve the accuracy of subsequent star map recognition.Finally,an improved star point extraction method is proposed.The improved Sobel operator is combined with the region growth method for rough star point extraction,and the Gaussian surface fitting method is used for centroid positioning to obtain centroid coordinates.Through simulation experiments on different extraction methods,comparative analysis of the error of star point extraction verifies the effectiveness of the improved algorithm.