作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2010, Vol. 36 ›› Issue (5): 168-169,. doi: 10.3969/j.issn.1000-3428.2010.05.061

• 人工智能及识别技术 • 上一篇    下一篇

基于DSP的实时目标特征提取算法

邓志云,宋志刚,王 荣   

  1. (中国空空导弹研究院,洛阳 471009)
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2010-03-05 发布日期:2010-03-05

Feature Extraction Algorithm for Real-time Target Based on DSP

DENG Zhi-yun, SONG Zhi-gang, WANG Rong   

  1. (China Airborne Missile Academy, Luoyang 471009)
  • Received:1900-01-01 Revised:1900-01-01 Online:2010-03-05 Published:2010-03-05

摘要: 成像制导武器系统进行多目标实时跟踪时,必须具备实时目标特征提取能力。实现一种基于TMS320C64x DSP平台的实时目标特征提取算法。针对C64x DSP的增强存储器访问特性,通过多像素信息的同步提取和像素归并方法,对算法进行优化。实验结果表明,优化后算法的执行效率和实时性较高。

关键词: 特征提取, 像素归并, 二值图像, 标记算法, 实时

Abstract: Imaging-guided weapon system must have the capability of real-time extraction for target feature during multi-target real-time tracking. This paper realizes a real-time extraction algorithm for target feature based on TMS320C64x DSP platform. According to the enhanced memory access characteristics of C64x DSP, the algorithm is optimized by using multi-pixel information synchronization extraction and pixel merging approach. Experimental results show that the optimized algorithm has high implementation efficiency and high real-time performance.

Key words: feature extraction, pixel merging, binary image, label algorithm, real-time

中图分类号: