Space Target Attitude Measurement Based On Image Processing

Space target's 3-D gesture is a hot issue in the field of aviation and military, Object's attitude mainly includes pitch angle, yaw angle and rolling angle. Currently, the research and applications in the filed is very popular. Using the method of binocular stereo vision measurement has important two important characteristics: non-contact measurement and high accuracy. So, the measurement system in the field of space target attitude measurement has been used widely.In this paper, space target attitude measurement is based on image processing. The moving target gesture measurement based on image processing and binocular stereo vision,which is under a single background. The implement mainly includes use binocular stereo vision imaging system to shoot the target images, get internal and external parameters from the camera calibration technique, the feature information of the image obtained from the image processing technology, finally, establish the mathematical model of attitude measurement algorithm and calculate the target angle at any time. The main work in the paper is divided into the following sections:First of all, we studied the camera calibration techniques: in order to complete the calibration of the two binocular stereo vision cameras,study the binocular stereo vision system, perspective projection model of camera, coordinate system of camera and the transformation between them. Considering the problem of space object pose measurement in this paper require an accurate results calculation. We mainly studied the principle and implementation process of Zhang plane template calibration method. The precision of calibration result is very high.Secondly, we studied the image pretreatment of the space target attitude measurement.According to the images characteristic information which is get from the binocular camera and the purpose to be completed. This paper mainly studies the image smoothing, threshold segmentation, edge detection, edge thinning and other common methods of image preprocessing. Through the experiment contrast, select the Gaussian filter to realize space target image smooth, ostu to realize the threshold segmentation, Canny operator to realize the edge detection and template method to realize refined edge. These are prepared for the subsequent calculation of the target attitude measurement.Third, attitude determination algorithm of the space axisymmetric target is a research focus of this article: in this paper, the target image is space axisymmetric bombs model, we first use Hough transform to extract two straight edges of the target image. On the basis of the central axis line extracted, adopting the method of the central axis line completed the pitch angle and yaw angle calculate of the space target. On the basis of the feature point extraction,spiral marked by pasting on the basis of the feature point extraction method based on the completion of the space target roll angle calculation. Through the method of the spiral by pasting marked to complete the roll angle calculation of space target. Large of experimental results demonstrate the effectiveness of the method.Finally,based on the algorithm, we build a space target attitude measurement system based on image processing in Visual C++ 6.0 platform. Given the interface operating instructions of the measurement system software simulation,and each operation simulation experiment shows the corresponding simulation results, including online image collection,camera calibration of the internal and external parameters, image preprocessing, feature information extraction and pitch angle, yaw angle and roll angle calculation function modules.Lastly, established the mathematical model of the space target attitude measurement system and finished the error analysis, given the factors that affect the system measurement results.Lay the foundation for the further improve measurement accuracy in the future.