Journal of Systems Engineering and Electronics ›› 2010, Vol. 32 ›› Issue (4): 851-853.

• 软件、算法与仿真 • 上一篇    下一篇

基于不变矩和禁忌搜索算法的图像识别方法

高勇,何刚,张晓晖   

  1. (西安理工大学自动化与信息工程学院电子工程系, 陕西 西安 710048 )
  • 出版日期:2010-04-23 发布日期:2010-01-03

Image recognition algorithm based on moment invariant and tabu search

GAO Yong, HE Gang, ZHANG Xiao-hui   

  1. (Dept. of Electronic Engineering, Faculty of Automation and Information Engineering, Xi’an Univ. of Technology, Xi’an 710048, China)
  • Online:2010-04-23 Published:2010-01-03

摘要:

为了满足图像识别较高的识别率和较低的时间消耗两方面的要求,提出了一种不变矩和禁忌搜索结合算法,即以不变矩作为特征参数,以禁忌搜索算法加快图像数据检测。在最佳的迭代步长下,不变矩和禁忌搜索结合算法在平均识别率、消耗时间、平均重复个数等参数方面,要明显优于一般不变矩算法。不变矩和禁忌搜索结合算法兼顾了识别率和时间消耗两个方面的因素,可以应用于一些高速的图像识别场合。

Abstract:

In order to meet the requirements of high recognition rate and low time consumption, an algorithm, the integration of moment invariant and tabu search, is presented, which regards the moment invariant as parameters and speads up the detection of image data by a tabu search method. Compared with the common moment invariant method, the integrative algorithm of moment invariant and tabu search is obviously better in the avarage recognition rate, average time consumption and average redundant recognition rate on condition that the best iterative step size. The proposed algorithm covers the situations of both recognition rate and time consumption and can be applied to high-speed image recognition occasions.