运动视觉跟踪的三维虚拟平台设计与实现

摘要

由于运动视觉跟踪对象随机性和多样性的特点，使得传统运动视觉跟踪的三维虚拟平台无法准确还原运动物体跟踪画面，决策误差较大。因此，构建还原精度与可应用性较高的运动视觉跟踪的三维虚拟平台。该平台由运动视觉跟踪监测模块、视频处理显示模块、运动视觉三维转换模块和计算机组成。利用运动视觉跟踪检测模块中的摄像机模拟人眼，进行运动视觉跟踪，将其转换成三维坐标视频文件。视频处理显示模块进行三维坐标视频文件的处理和显示工作，将视频文件的编码传输到运动视觉三维转换模块。运动视觉三维转换模块将视频编码转换成三维虚拟图像，为使用者提供决策信息。软件给出了非线性跟踪下，三维虚拟画面的生成流程图，以及AM209mm视频处理器修正运动视觉跟踪画面的算法语言。实验结果表明，所设计平台具有较高的还原精度与可应用性。%The traditional 3D virtual platform for the moving vision tracking can't accurately restore the pictures from moving object tracking,may make a bigger decision error because of the randomness and diversity characteristics of moving vision tracking object. Therefore,the 3D moving visual tracking virtual platform with high restoring precision and applicability was built. The platform is composed of moving visual tracking monitoring module,video processing display module,moving visual 3D conver?sion module and computer. Of vision tracking detection module in The camera in moving the visual tracking monitoring module is utilized to simulate human eye to perform the moving visual tracking,and convert the tracking results into a 3D coordinate video files. The video processing display module is used to process and display for 3D coordinate video files,and transmit the codes of video files to the moving visual 3D conversion module. The moving visual 3D conversion module is adopted to convert the video codes into 3D virtual images for providing the user with decision?making information. The 3D flow chart of the virtual image generation in the nonlinear tracking is given by the software. The algorithmic language for correcting the moving visual tracking pictures by AM209mm video processor is offered. The experimental result shows that the platform has high restoring pre?cision and applicability.