计算机科学 ›› 2017, Vol. 44 ›› Issue (Z6): 169-172.doi: 10.11896/j.issn.1002-137X.2017.6A.039

• 模式识别与图像处理 • 上一篇    下一篇

基本图像处理算法的优化过程研究

徐启航,游安清,马社,崔云俊   

  1. 中国工程物理研究院应用电子学研究所 绵阳621900,中国工程物理研究院应用电子学研究所 绵阳621900;高功率微波技术国防科技重点实验室 绵阳621900,中国工程物理研究院应用电子学研究所 绵阳621900,中国工程物理研究院应用电子学研究所 绵阳621900
  • 出版日期:2017-12-01 发布日期:2018-12-01
  • 基金资助:
    本文受国防微波重点实验室基金(2015HPM-11)资助

Study on Optimizations of Basic Image Processing Algorithm

XU Qi-hang, YOU An-qing, MA She and CUI Yun-jun   

  • Online:2017-12-01 Published:2018-12-01

摘要: 针对实时、高效的图像处理任务的实现,以视频图像中基于模板匹配的运动目标开环跟踪算法为例,对其基于Matlab原型算法的跟踪性能进行评估,具体介绍了对此算法的多级优化过程。从Matlab原型算法开始,主要 从以下两方面进行优化:在提高实时处理速度方面,采用C语言提速、乘法提速、Release提速、合并运算、CUDA架构提速等10级以上的优化策略;在提高正确率方面,采用简单的多模板策略。测试结果表明,算法速度提高了200多倍,最终达到30Hz的实时处理速度,并且大幅提高了跟踪正确率。

关键词: 模板匹配,提速,多模板策略,目标跟踪,图像处理,CUDA架构

Abstract: To provide useful references for the implementation of real-time and excellent image processing task,taking the open loop tracking algorithm based on template matching in video image as an example,the tracking performance of the algorithm based on MATLAB prototype algorithm was evaluated,and the multi-level optimization process was pre-sented.Starting with an MATLAB prototype algorithm,we began to optimize mainly in the two aspects as below.In improving real-time processing speed,more than 10 levels of optimizations are applied to speed up the algorithm,including C language speeding,multiplication speeding,release speeding,merge operation,CUDA speeding and so on.In terms of improving the correct rate,simple multi-pattern strategy is used.Testing results indicate that the algorithm reaches performance of 30Hz real-time image process and tracking rate of the algorithm is also promoted greatly.

Key words: Pattern match,Speed promotion,Multi-pattern strategy,Targets tracking,Image processing,CUDA architecture

[1] 焦波.面向智能视频监控的运动目标检测与跟踪方法研究[D].长沙:国防科技大学,2009.
[2] 徐治非.视频监控中运动目标检测与跟踪方法研究[D].上海:上海交通大学,2009.
[3] 傅荟璇.运动目标识别与光电跟踪定位技术研究[D].哈尔滨:哈尔滨工程大学,2009.
[4] 郭伟.复杂背景下红外目标检测与跟踪[D].西安:西安电子科技大学,2008.
[5] 李位星,范瑞霞.基于DSP的运动目标跟踪系统[J].自动化技术与应用,2004,23(4):46-49.
[6] 梁路宏,艾海舟,肖习攀,等.基于模板匹配与支持矢量机的人脸检测[J].计算机学报,2002,25(1):22-29.
[7] CUDA Reference Manual(Version 2.0)[M].Santa Clara,Nvidia Corporation,2008.
[8] CUDA Programming Guid(Version 2.0) [M].Santa Clara,Nvidia Corporation,2008.
[9] 邓浩,杨菲,潘旭东,等.利用CUDA实现上位机的实时目标跟踪[J].信息与电子工程,2010,8(3):368-371.
[10] 邵平,杨路明,黄海滨,等.基于积分图像的快速模板匹配[J].计算机科学,2006,33(12):225-229.
[11] 魏国剑,侯志强,李武,等.融合光流检测与模板匹配的目标跟踪算法[J].计算机应用研究,2014(11):3498-3501.
[12] 武亮,孙秦.有限元的OpenMp并行计算技术[J].航空计算技术,2013,43(5):56-60.
[13] ISARD M,BLAEK A.CONDENSATION—Conditional Density Propagation for Visual Tracking[J].International Journal of Computer Vision,1998,29(1):5-28.

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!