SIA OpenIR  > 智能产线与系统研究室
电气符号识别的HOG方法
Alternative TitleResearch on Electrical Symbols Recognition of HOG
刘剑; 龚志恒; 高恩阳; 刘亚楠
Department装备制造技术研究室
Source Publication沈阳建筑大学学报(自然科学版)
ISSN2095-1922
2013
Volume29Issue:3Pages:571-576
Contribution Rank3
Funding Organization国家自然科学基金项目(61272253);; 住房和城乡建设部科技项目(2010-K9-22)
Keyword电气符号识别 Hog特征 支持向量机 梯度直方图
Abstract目的为解决电气符号的大小、图纸背景的模糊、电气符号的旋转角度等各种干扰因素对计算机识别电气图纸造成的误差问题.方法笔者提出了一种基于HOG的电气符号识别方法.建立电气符号训练集,提取电气符号图像的HOG特征,计算出梯度方向向量个数加权图;使用这些HOG特征和分类信息对支持向量机进行训练;利用支持向量机进行识别.结果HOG算法对电气符号的识别率达到92.5%,与SIFT算法比较,识别效果更为准确.结论所提出的HOG算法克服了外界干扰因素对电气符号识别的影响,提高了识别的准确率,具有良好的检测效果,为将HOG算法应用到其他领域奠定理论基础.
Other AbstractElectrical symbols recognition is one of the hot spots in the field of computer vision and electrical design recently. Various factors will bring errors on computer recognition of electrical drawings, such as size of the electrical symbols, drawings of background fuzzy, rotation of the electrical symbols etc.. There are many electrical symbols recognition methods, but the problems are solved hardly. In order to overcome the external interference factors and improve recognition accuracy, an electrical symbol recognition method based on HOG is presented in the paper. Firstly, the electrical symbol training sets were established, and symbols of HOG feature were extracted. Secondly, support vector machine was trained by mean of these characteristics of HOG and classification information. Finally, symbols were recognized by support vector machines. Con- ducted different experiment of complex background research of electrical symbols and experimental study on the algorithm of the algorithm of SIFT and HOG, the results of experiments show that the proposed algorithm of HOG has good detection after experiment.
Language中文
Document Type期刊论文
Identifierhttp://ir.sia.cn/handle/173321/12452
Collection智能产线与系统研究室
Affiliation1.沈阳建筑大学信息与控制工程学院
2.东北大学信息科学与工程学院
3.中国科学院沈阳自动化研究所
Recommended Citation
GB/T 7714
刘剑,龚志恒,高恩阳,等. 电气符号识别的HOG方法[J]. 沈阳建筑大学学报(自然科学版),2013,29(3):571-576.
APA 刘剑,龚志恒,高恩阳,&刘亚楠.(2013).电气符号识别的HOG方法.沈阳建筑大学学报(自然科学版),29(3),571-576.
MLA 刘剑,et al."电气符号识别的HOG方法".沈阳建筑大学学报(自然科学版) 29.3(2013):571-576.
Files in This Item: Download All
File Name/Size DocType Version Access License
电气符号识别的HOG方法.pdf(517KB) 开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[刘剑]'s Articles
[龚志恒]'s Articles
[高恩阳]'s Articles
Baidu academic
Similar articles in Baidu academic
[刘剑]'s Articles
[龚志恒]'s Articles
[高恩阳]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[刘剑]'s Articles
[龚志恒]'s Articles
[高恩阳]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 电气符号识别的HOG方法.pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.