期刊文献+

基于FPGA并行处理SIFT算法特征点检测 被引量:4

Realization of SIFT Feature Detection Parallel Implementation Based on FPGA
下载PDF
导出
摘要 由于SIFT算法计算量大,以至于不能够应用于实时性要求高的领域。为了使SIFT算法满足实时性的需求,详细阐述了SIFT算法中特征点检测与梯度计算硬件实现过程和系统架构,并且对SIFT算法进行相应的改进,使其适合FPGA并行处理特性。最后利用Altera公司Cyclone III系列芯片的硬件平台测试SIFT算法硬件部分的计算时间和匹配性能,并且与DSP实现的SIFT算法进行性能上的比较。结果表明,FPGA实现SIFT算法具有实时性高,匹配性能好的特点。 Due to large calculated amount of SIFT algorithm, it is not used in real-time field. In order to the real-time demand for SIFT algorithm. Hardware realization and system architecture that characteristic points detection and calculating pixel' s gradient calculation in SIFT algorithm are realized by FPGA is introduced in this paper. And SIFT algorithm is optimized for the parallel processing characteristic of FPGA. Calculating time and matching performance can be test by FPGA IC chip of the Altera Cyclone III series and matching performance can be compared SIFT in FPGA with SIFT in DSP. Finally, FPGA implementation of SIFT algorithm have characteristic of high real-time and matching performance.
出处 《电视技术》 北大核心 2012年第23期188-192,共5页 Video Engineering
关键词 SIFT FPGA 特征检测 尺度空间 SIFT FPGA feature detection scale space
  • 相关文献

参考文献7

二级参考文献35

共引文献78

同被引文献39

  • 1李海巍,杨敏华,刘益世.基于中值滤波和小波变换的遥感影像去云[J].测绘与空间地理信息,2012,35(7):49-51. 被引量:7
  • 2HARRIS C, STEPHENS M J. A combined comer and edge detector [C]//Proc. 4th Alvey Vision Conference. Manchester, UK: [ s. n. ], 1988:147-151.
  • 3CHANDAR1A J,THOMAS G, BARTCZAKB, et al. Real-time camera tracking in the MATRIS project [ J ]. SMPTE Motion Imaging Journal, 2007,116(7) :266-271.
  • 4KITCHEN L, ROSENFELD A. Gray-level corner detection [ J ]. Pattern Recognition Letters, 1982 ( 1 ) :95-102.
  • 5BAR-SHALOM Y,LI X R. Estimation and tranking: principles, tech- niques, and software[ M 1. Norwell : Artech House, 1993.
  • 6FELSBERG M,GRANLUND G H. POI detection using channel cluste- ring and the 2D energy tensor [ C ]//Proc. 26th DAGM Symposiumg. Tubingen, Germany : Springer Berlin ,2004 : 103-110.
  • 7Euclidean distance [ EB/OL ]. [ 2012-09-121. http ://en. wikipedia. org/wiki/Euclidean_distance.
  • 8MARSHALL D. Nearest neighbor searching in high dimensional metric space. [ EB/OL ]. [ 2012-09 - 12 ]. http ://escience. anu. edu. Au/pro- ject/O6S1/DavidMarshall, 2006-07-21 /2009-12.
  • 9BEIS J S, LOWE D G. Shape indexing using approximate nearestneighbor search in high-dimensional spaces [ C ]//Prec. CVPR 1997. IS. 1. ] :IEEE Press ,1997 :1000-1006.
  • 10LOWE D G. Object recognition from local scale-invariant features [C]//Proc. ICCV. Kerkyra, Greece:[s.n.], 1999,7(2) : 1150-1157.

引证文献4

二级引证文献7

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部