[1]翟传敏,杜吉祥,黄 飞.基于径向基概率神经网络的工程图纸图形符号识别[J].智能系统学报,2006,1(1):88-91.
ZHAI Chuan-min,DU Ji-xiang,HUANG Fei.Graphic symbol recognition of engineering drawings based on radial basis probabilistic neural networks[J].CAAI Transactions on Intelligent Systems,2006,1(1):88-91.
点击复制
ZHAI Chuan-min,DU Ji-xiang,HUANG Fei.Graphic symbol recognition of engineering drawings based on radial basis probabilistic neural networks[J].CAAI Transactions on Intelligent Systems,2006,1(1):88-91.
基于径向基概率神经网络的工程图纸图形符号识别
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
1
期数:
2006年第1期
页码:
88-91
栏目:
学术论文—机器感知与模式识别
出版日期:
2006-03-25
- Title:
- Graphic symbol recognition of engineering drawings based on radial basis probabilistic neural networks
- 文章编号:
- 1673-4785(2006)01-0088-04
- Keywords:
- radial basis probabilistic neural network; graphicsy mbol; engineering drawings recognition
- 分类号:
- TP31
- 文献标志码:
- A
- 摘要:
- 基于径向基概率神经网络，提出一种扫描工程图纸图像分割后的图形符号识别方法.针对已分割的扫描工程图纸图形符号图像，首先进行二值化处理，然后对二值图形符号图像进行Hu不变矩特征提取，再使用一种新型的径向基概率神经网络进行分类，从而实现图像识别.为加快径向基概率神经网络的收敛速度，采用递归最小二乘算法进行训练.实验结果表明，径向基概率神经网络在识别性能与速度等方面非常适合于工程图纸的图形符号识别.
- Abstract:
- A novel graphic symbol recognition approach of engineering drawings based on radial basis probabilistic neural network s (RBPNN) is proposed. The Hu invariant moment method is applied to extract the shape features of the segmented graphic symbol image of scanned engineering draw ings. The experimental results show that the RBPNN achieves a higher recognition rate and better classification efficiency with respect to radial basis function neural networks (RBFNN) and multi-layer perceptron networks (MLPN) for the gra phic symbol recognition task.
- 参考文献/References:
-
［1］董玉德,赵??? 韩,王??? 平，等.工程图纸识别与理解的研究现状分析［J］. 合肥工业大学学报(自然科学版), 2005, 28(1):29-33.
DONG Yude, ZHAO Han, WANG Ping, et al. Analysis of research status of enginee rin g drawings recognition and interpretation ［J］. Journal of Hefei University of Technology, 2005, 28(1):29-33.
［2］黄德双. 神经网络模式识别系统理论［M］. 北京:电子工业出版社, 1996.
HUANG Deshuang. Systematic theory of neural networks for pattern recognitio n ［M］. Beijing: Publishing House of Electronic Industry of China, 1996.
［3］HUANG Deshuang. Radial basis probabilistic neural networks: model a nd application ［J］. International Journal of Pattern Recognition and Artificia l Intelligence, 1999, 13(7): 1083-1101.
［4］赵温波, 杨鹭怡, 王立明.径向基概率神经网络的混合结构优化算法［J］. 系统仿真学报, 2004, 16(10): 2175-2180.
?ZHAO Wenbo, YANG Luyi, WANG Liming. The hybrid structure optimization algo rithms of radial basis probabilistic neural networks［J］. Journal of System Sim ul ation, 2004, 16(10): 2175-2180.
［5］HUANG Deshuang, ZHAO Wenbo. Determining the centers of radial basi s probabilistic neural networks by recursive orthogonal least square algorithms ［J］. Applied Mathematics and Computation, 2005, 162: 461-473.
［6］梁德本, 叶玉驹. 机械制图手册3版［M］. 北京: 机械工业出版社, 200 5.
LIANG Deben, YE Yuju. Manual of mechanical drawing:3rd ed［M］. B eijing: Publishing House of Mechanical Industry of China, 2005.
［7］冈萨雷斯. 数字图像处理： 第2版［M］.阮秋琦,阮宇智，等译. 北京: 电子工业出版社, 2003.
GONZALEZ C. Digital image processing:2nd ed ［M］.Beijing: Pub lishing House of Electronic Industry of China, 2003.
［8］CHEN C C. Improved moment invariants for shape discriminat ion ［J］. Pattern Recognition, 1993, 26(5):683-686.
备注/Memo
收稿日期：2006-02-15.
基金项目：国家自然科学基金资助项目(60405002)；合肥学院自然科学研究基金资助项目(05ky013zr).
作者简介：
翟传敏，女，1977年生，讲师，2002年在合肥工业大学获工学硕士学位.主要从事模式识别、机械CAD相关研究工作.
杜吉祥，男，1977年生，博士后，2005年在中国科学技术大学获工学博士学位. 主要研究方向为模式识别与图像处理等.已在国际会议和国际杂志上发表论文10余篇，大部分被SCI、EI收录.
黄??? 飞，男，1978年生，讲师，2002年在合肥工业大学获工学硕士学位.主要从事CAD /CAM、测控技术及仪器科学技术的研究.
更新日期/Last Update:
2009-04-07