Research And Implementation Of Key Technologies For Mobile Multi-target Wireless Laser Range

With the continuous development of computer technology and information technology,aiming at the needs of laser simulation shooting target system in various occasions such as military training for college or university freshmen,social practice or military training for primary and secondary school,grassroots units army and the police department,a variety of national defense education base and patriotic education venues,military training of army and military theme entertainment venues,a mobile multi-target wireless laser range system is designed based on image processing.It is difficult to popularize the shooting training because of the cost,training conditions,regulations and safety hazards.However,the system in this thesis has the advantages of low cost,safety and practicality,and it has good application prospect and meets the requirements of accuracy and real-time shooting.The main search contents of this thesis are as follows:(1)The status quo of the laser shooting system and its related techniques are analyzed.The status quo of the laser shooting system and its related technologies are analyzed.The system overall framework is designed based on the characteristics and system requirements of the simulated laser shooting.The ideas and workflow of each part and key technologies such as mobile multi-target positioning are also described.(2)According to the requirements of multi-target accurate positioning in shooting training,a multi-target positioning method based on FAsT-Match algorithm is studied.The algorithm can adapt to the two-dimensional affine transformed image such as translation,scaling and rotation,and is not sensitive to the illumination of the image,and has good anti-interference ability to the background noise.Then the fuzzy c-means clustering algorithm is used to achieve the objective of targeting multiple targets.(3)For moving targets,the camera captures a sequence of image frames.Aiming at the shooting training in complex environment,the multi-target location method based on convolution neural network is studied in combination with the idea of deep learning.A set of target data sets is set up to train a convolution neural network with eight layers of structure.Then the related edge information in the each frame image captured from one shoot is detected,and the Edge Boxes algorithm is used to find the area that may be the target based on a threshold.The region with no target is filtered through the convolutional neural network.And the region with no target is filtered through the convolutional neural network.Finally,the optimal result is obtained by removing redundant candidate regions according to the non-maximum suppression method.(4)The mobile multi target wireless laser range system is implemented,and the application of the system is given.And the system works well.