毕业论文

打赏
当前位置: 毕业论文 > 自动化 >

MATALB视频目标跟踪性能评估软件设计

时间:2018-03-28 11:45来源:毕业论文
借助MATALB GUI平台设计了一款评估软件。首先,对评估指标进行了选取,并设计了相关算法,根据评估要求设计了七个系统功能模块,并利用GUI界面展示出来

摘要视频跟踪广泛应用于生产实践的各个领域,跟踪算法大量涌现,但跟踪性能评估软件却较为缺失,评估软件的稀缺阻碍了跟踪算法的进一步发展。本文针对这种情况借助MATALB GUI平台设计了一款评估软件。首先,对评估指标进行了选取,并设计了相关算法,根据评估要求设计了七个系统功能模块,并利用GUI界面展示出来,实验证明指标之间相互补充,能较好地反应系统性能,系统功能较为完善;其次,给出了使用该软件的三种评估操作方法,结合五种典型跟踪算法和五种典型异常情况,对软件进行仿真与结果分析,仿真结果表明该软件评估方式全面,评估效果可靠。
关键词  目标跟踪  性能评估指标  GUI界面设计  评估方式   20283
毕业论文设计说明书(论文)外文摘要
Title   Software Design in Performance Evaluation of   Video Object Tracking
Abstract
Video tracking is widely used in various fields of production practice with the emergence of tracking algorithms in large numbers.However,there is a lack of  tracking performance evaluation software,hindering the further development of tracking algorithms.In view of this situation,this paper designed a piece of evaluation software with the platform of MATLAB GUI.First of all, the evaluation indicators were selected, and the related algorithms were designed.Seven system function modules were designed according to the assessment requirements,displayed with the GUI interface.It was proved that there were mutual complements between experiment indexes by the experiment.The indexes can reflect the system performance accurately,while the system functions were complete.Secondly, three kinds of evaluation methods were given to operate the software.The software was simulated with the combination of evaluation methods,five typical tracking algorithms and five typical abnormal situations.The results were analysed,and they were found that the assessment methods were comprehensive and the evaluation effect was reliable.
 Keywords  object tracking, performance evaluation index, GUI based interface design, assessment method
目次   
1  引言    1
1.1  研究背景和意义    1
1.2  研究现状    1
1.3  本文主要工作    2
2  评估指标及算法介绍    3
2.1  评估指标选取要求    3
2.2  评估指标的选定及分析    4
2.3  跟踪算法简介    8
3  软件设计与实现    11
3.1  开发环境介绍    11
3.2  系统功能介绍    11
3.3  GUI界面布局    16
4  评估方式及测试数据集    17
4.1  评估方式简介    17
4.2  测试数据集简介    18
5  算法评估实验与结果分析    21
5.1  单次跟踪评估下的算法性能    21
5.2  鲁棒性综合评估下的算法性能    22
5.3  异常情况下的算法评估及分析    29
5.4 结果分析与讨论    34
结论    36
致谢    37
参考文献    38
1  引言
1.1  研究背景和意义
    视频目标跟踪是计算机视觉中至关重要的一项技术,它是多相关领域先进技术的利用、融合、延伸与发展。随着计算机技术的高速发展及日益增长的民用军用需要,越来越多的跟踪算法 涌现出来,如光流法、帧间差分法、粒子滤波跟踪算法等等。跟踪算法各有其优势及缺陷,而对于不同的场景需要及情况需要,算法的选择亦应因地制宜。综合考虑后可知,当前视频目标跟踪领域急需性能评估软件,来对现今纷繁复杂的跟踪算法进行度量、评估和分类,视频目标跟踪性能评估软件的设计与研究显得尤为必要而迫切。 MATALB视频目标跟踪性能评估软件设计:http://www.youerw.com/zidonghua/lunwen_12004.html
------分隔线----------------------------
推荐内容