Research On Information Extraction And Inversion Of Fragment Field Based On Image Processing Technology

Modern war has put forward higher requirements on the killing power and damage effect of warhead.All countries in the world are actively researching and developing new warhead technologies with various damage mechanisms in order to improve the high-efficiency damage ability of ammunition and missile.Warhead is an important part of missile weapon system and the main component to complete the mission of warhead.According to the characteristics and mechanism of damage formation,it can be divided into anti-personnel warhead,blasting warhead,penetrating warhead,shaped warhead.Anti-personnel warhead mainly presents fragment damage,and the number of fragment perforation,velocity,scattering angle,scattering direction angle,density distribution and fragment perforation area are important indexes to judge the effectiveness of anti-personnel warhead.In the process of warhead development,the fragmentation parameters are generally measured and tested in static detonation test in the shooting range,which provides direct basis for the warhead evaluation,design verification and final acceptance.The fragment dispersion characteristics of anti-personnel warhead mainly depend on manual measurement for statistical analysis and calculation,and the measurement parameters presents large error and low efficiency.In this paper,the digital close-range photography and image processing were adopted to realize the fast and accurate acquisition and 3D inversion of the dispersion characteristics of explosive fragment.It made up for the shortage of manual methods and the rapid assessment of the warhead power to provided testing methods comprehensively and accurately.The research content of this paper mainly includes:（1）The method of image acquisition of close-range photogrammetric was studied.The information extraction and 3D inversion system of fragment field were established based on image processing technology.Aiming at the problem that the 3D splicing feature of the target plate layout scene was not obvious before explosion and the splicing feature design of the explosion scene was designed by optimizing the splicing feature design.According to the invariable shape characteristics of the target plate and explosion center,and the reflective mark and coding mark were proposed to be embedded in the target plate to design the splicing feature of the explosion scene.The sequence images of the scene before explosion were obtained by collecting the net shape design,and provided effective image information for the reconstruction of the scene of target plate layout.（2）Explosion scene sequence image processing and multi-feature fusion extraction algorithm.Multi-feature algorithms such as threshold segmentation of maximum inter-class variance,regional connectivity labeling,affine transformation decoding and gradient amplitude mean were used to calculate the coordinate of the target sequence image.Based on planar homographic matrix of camera parameters and distortion calibration model,using encoding matching relative orientation and benchmark ruler absolute orientation,implement resection of pose estimation.Image triangulation was presented and overall adjustment optimization method to realize single point symbol and space of the target plate setting scene3 D inversion.（3）3D inversion and calculation of flight parameters of explosion fragment field.The position and pose of target plate change after explosion,which bring difficulties to the inversion and reconstruction of 3D measurement of fragment perforation and pits.Firstly,the target plate distortion correction was realized based on collinearity equation model,and adaptive median filter algorithm was adopted to improve the filtering.The piecewise linear gray scale transformation method was proposed to enhance the image contrast analysis.Combined with Canny edge detection algorithm,active contour energy function edge extraction algorithm,Gabor wavelet transform and gray co-occurrence matrix texture feature extraction algorithm,interference target algorithm based on mathematical morphology,fusion feature target recognition algorithm,achieve fast and accurate segmentation of fragmentation perforation and pits.The identification rate was 100% when fragments larger than 3mm.Secondly,the center coordinate position and area were calculated by using the connected region based target annotation and target feature gray weighted center of gravity method.A new method was proposed to calculate the coordinates of the center and fragments area of irregular fragments by weighted center of gravity method of target feature,which improved the calculation accuracy of irregular fragments by 20% and the calculation efficiency by more than 10 times.Then combined with the relative position and pose relationship of target plate layout scene established before explosion.The velocity adjustment method was used to study the 3D positions of the perforation and pit relative to the explosion center,and the 3D measurement inversion was realized.Finally,according to the corresponding relationship between 3D spherical projection and target plate,the relevant data processing software was formed,and the parameters of the rupture field were calculated automatically.At the same time,the measurement accuracy was analyzed and the field test was verified.