A Rapid Detection Technology Research Of Moving Spatial Objects

Space debris refers to non-functional man-made objects left on Earth orbit or reentered into the atmosphere by human activities in space.It mainly includes discarded rocket and rocket bodies,satellite bodies,and fragments generated by collisions between space objects.It is the main destroyer of space environment.In recent years,the amount of space debris is increasing rapidly and the damage to the space environment is increasing day by day.The average relative speed between these fragments and the spacecraft in orbit is about 10km/s,and the extremely high pressure generated during the impact causes serious damage to the spacecraft functions,and even leads to the complete disintegration of the spacecraft.Space debris detection is an important part of space detection,which is of great significance to the construction of spatial object monitoring system.Ground-based optical telescope is one of the main means of space exploration at present,which has certain advantages in terms of observation cost and technical use.Taking photometric images through ground-based optical telescopes is an important means to analyze,detect and track spatial objects.In astronomical observation,spatial objects must be accurately located in order to calculate its orbital parameters which is to avoid collision by tracking and forecasting.Spatial object detection is directly related to cataloging,tracking and spacecraft operation.It usually takes two steps from taking an image from a ground-based optical telescope to detecting and extracting a spatial object from the image.The first step is photometric image processing,which mainly uses image processing technology to extract spatial objects from photometric images.The second step is spatial object detection,which uses the spatial object detection method to detect the moving spatial object from the above spatial objects,and then completes the spatial object detection.Therefore,a spatial object detection method based on 3D data is proposed by combining photometric image processing and spatial object detection to recognize spatial moving spatial objects in images.The structure of this paper is as follows:Firstly,the method of photometric image processing is introduced.This paper presents the process of traditional photometric image processing:updating file header,image purification,image correction,source extraction,astronomical positioning,flux calibration,aperture photometry.Combined with image transformation and source extraction,the principle and function of these processes are introduced in detail.In the data processing before the spatial object detection,this paper introduces the image processing process and the function of each step.Secondly,the moving spatial object detection method is introduced.This paper first introduces the traditional spatial moving spatial object detection techniques:optical flow method,frame difference method and background difference method.On this basis,combined with the requirements of moving weak spatial object detection in space,the spatial object detection is transferred from the two-dimensional plane to the three-dimensional point cloud,use the linear recognition method-three-dimensional Hough transform to accurately and quickly detect the trajectory of the spatial object,complete the recognition and extraction of the spatial object in order to develop a spatial object detection method based on the three-dimensional data.Finally,the spatial object detection method based on 3D data is tested by using simulation data and measured data.By analyzing the simulation image recognition results,it is concluded that the spatial object recognition rate increases with the increase of spatial object signal-to-noise ratio and the number of detection group images.Using measured images to test,we can extract weak spatial objects with a signal-to-noise ratio below 5,and the extraction effect is good for spatial objects with a signal-to-noise ratio higher than 5.Asteroids can be identified using asteroid tracking images and star tracking images.The spatial object recognition speed of the measured image is tested,which can meet the requirements of real-time image processing.