期刊文献+

生猪虹膜快速定位算法研究

Research on fast localization algorithm of pig iris
下载PDF
导出
摘要 虹膜定位是实现虹膜识别的前提与关键,当前主流虹膜采集设备普遍价格昂贵,且现有虹膜定位算法应用于猪眼虹膜定位时存在精度低、耗时长等问题。本文提出一种基于低成本采集设备的生猪虹膜快速高精度定位算法。首先对猪眼虹膜图像去噪和二值化,再对Canny算法进行改进,用于提取猪眼虹膜图像边缘信息,然后分别使用改进的Hough圆变换和迭代加权最小二乘法定位虹膜内、外边缘,进而完成生猪虹膜定位。对3100张猪眼虹膜图像进行试验,结果表明,虹膜的正确检测率为96.19%,平均定位时间为342.50 ms。与DAUGMAN、WILDES和单位扇环灰度算法相比,正确检测率分别提升了10.49、8.67、3.61百分点,平均定位时间分别减少了70.90、81.50、98.00 ms。本文方法能在保证生猪虹膜正确检测率的基础上缩短定位时间,有效改善传统虹膜定位算法应用于猪眼虹膜时效果较差的问题,能为后续基于虹膜的生猪身份识别奠定基础。 Iris localization is the prerequisite and key link to realize iris recognition,the current mainstream iris capture devices are generally expensive,and the existing iris localization algorithm have problems such as low accuracy and long time consumption when applied to pig eye iris localization,so this paper proposes a fast and high accuracy iris localization algorithm for pigs based on low-cost acquisition equipment.Firstly,the pig iris image was denoised and binarized,then the Canny algorithm was improved to extract the edge information of the pig iris image,and then the inner and outer edges of the iris were located using the improved Hough circle transform and iterative reweighted least square method,so as to complete the localization of the pig iris.Experiments on 3100 pig eye iris images showed that the accuracy rate of the iris was 96.19%and the average localization time was 342.50 ms.Compared with the DAUGMAN,WILDES and unit sector ring grey scale algorithms,the accuracy rate was improved by 10.49 percentage points,8.67 percentage points and 3.61 percentage points,respectively,and the average localization time was reduced by 70.90,81.50,98.00 ms.This method could shorten the localization time while ensuring accuracy rate,and effectively improve the problem of poor effect of traditional iris positioning algorithm applied to pig iris,so as to lay an important foundation for the subsequent iris-based pig identification.
作者 刘伟东 周素茵 徐爱俊 叶俊华 LIU Weidong;ZHOU Suyin;XU Aijun;YE Junhua(College of Mathematics and Computer Science,Zhejiang A&F University,Hangzhou 311300,China;College of Environment and Resources,Zhejiang A&F University,Hangzhou 311300,China)
出处 《浙江农业学报》 CSCD 北大核心 2024年第11期2617-2626,共10页 Acta Agriculturae Zhejiangensis
基金 浙江省“领雁”研发攻关计划(2022C02050) 浙江省“三农九方”科技协作计划(2022SNJF057)。
关键词 生猪 虹膜定位 CANNY算法 Hough圆变换 正确检测率 pig iris localization Canny algorithm Hough circle transform accuracy rate
  • 引文网络
  • 相关文献

参考文献19

二级参考文献306

共引文献243

;
使用帮助 返回顶部