基于支持向量机的轮胎标识点形状识别

基于支持向量机的轮胎标识点形状识别

摘要：轮胎标识点在轮胎分类中发挥着重要的作用，为了解决人工识别标识点#br# 工作强度大、误差较高等问题，提出了一种基于支持向量机的轮胎标识点形状识别算法。首#br# 先利用中值滤波对标识点图像进行降噪处理，根据标识点图像的灰度直方图对图像进行分割#br# 以去除背景，并利用Canny 边缘检测算子提取标识点边缘轮廓。然后，提取标识点轮廓的#br# 傅立叶描述子系数，最后利用支持向量机对标识点的傅立叶描述子系数进行分类以识别标识#br# 点的形状。实验结果表明，该算法能准确地识别出轮胎标识点的形状，提高了标识点识别效#br# 率。#br# 关键词：支持向量机；标识点；傅立叶描述子；形状识别

出版日期:2015-04-30 发布日期:2015-03-30

The Shape Recognition of Tyre Marking Points Based on Support Vector Machine

Abstract: The tyre marking points play an important role in the tyre classification. However,#br# manual recognition is laborious and is easily influenced by physiological factors. A method of#br# shape recognition of tyre marking points based on Support Vector Machine(SVM) is proposed in#br# this paper. Firstly, the image of tyre marking points will be denoised by median filter and#br# segmented based on the gray level histogram. Then the edge of tyre marking points will be#br# detected by Canny algorithm, and the detected edge will be represented by Fourier transform.#br# Lastly, those Fourier coefficients will be used as input to SVM classifier for distinguishing shapes#br# of tyre marking points. The experimental results show that the proposed method can achieve high#br# correct shape recognition rate and improve the efficiency of recognition of tyre marking points.#br# Key words: support vector machine; tyre marking points; Fourier descriptors; shape#br# recognition

Online:2015-04-30 Published:2015-03-30

摘要/Abstract

摘要： 轮胎标识点在轮胎分类中发挥着重要的作用，为了解决人工识别标识点
工作强度大、误差较高等问题，提出了一种基于支持向量机的轮胎标识点形状识别算法。首
先利用中值滤波对标识点图像进行降噪处理，根据标识点图像的灰度直方图对图像进行分割
以去除背景，并利用Canny 边缘检测算子提取标识点边缘轮廓。然后，提取标识点轮廓的
傅立叶描述子系数，最后利用支持向量机对标识点的傅立叶描述子系数进行分类以识别标识
点的形状。实验结果表明，该算法能准确地识别出轮胎标识点的形状，提高了标识点识别效
率。

关键词: 支持向量机, 标识点, 傅立叶描述子, 形状识别

Abstract: The tyre marking points play an important role in the tyre classification. However,
manual recognition is laborious and is easily influenced by physiological factors. A method of
shape recognition of tyre marking points based on Support Vector Machine(SVM) is proposed in
this paper. Firstly, the image of tyre marking points will be denoised by median filter and
segmented based on the gray level histogram. Then the edge of tyre marking points will be
detected by Canny algorithm, and the detected edge will be represented by Fourier transform.
Lastly, those Fourier coefficients will be used as input to SVM classifier for distinguishing shapes
of tyre marking points. The experimental results show that the proposed method can achieve high
correct shape recognition rate and improve the efficiency of recognition of tyre marking points.

Key words: support vector machine, tyre marking points, Fourier descriptors, shape
recognition

王勇，郭慧. 基于支持向量机的轮胎标识点形状识别[J]. 图学学报.

Wang Yong, Guo Hui. The Shape Recognition of Tyre Marking Points Based on Support Vector Machine[J]. Journal of Graphics.

[1]	黄杰, 魏欣, 杨子元, 闵卫东. 基于多维匹配距离融合的指节纹识别[J]. 图学学报, 2022, 43(2): 279-287.
[2]	张成 , 侯宇超 , 焦宇倩 , 白艳萍 , 李建军 . 基于三通道分离特征融合与支持向量机的混凝土图像分类研究[J]. 图学学报, 2021, 42(6): 917-923.
[3]	徐识溥, 李琳一, 贾金原, 王运圣, 刘畅, 刘勇, 马超. AR 辅助的基于 HOG-SVM 移动水稻病害智能分析与识别系统[J]. 图学学报, 2021, 42(3): 454-461.
[4]	李骏, 杨雅志 . 基于深度学习与有向无环图 SVM 的局部调整年龄估计[J]. 图学学报, 2021, 42(1): 1-7.
[5]	刘昶, 徐超远, 张鑫, 薛磊. 液晶字符识别的 CNN 和 SVM 组合分类器[J]. 图学学报, 2021, 42(1): 15-22.
[6]	曹鹏辉 , 吕书强 , 汪万福 , 高振华 , 侯妙乐 , . 基于光谱降维与 Hu 矩的壁画颜料层脱落区域提取方法 [J]. 图学学报, 2020, 41(6): 930-938.
[7]	石志良，张鹏飞，李晓垚. 基于 SIFT-SVM 的发动机主轴承盖识别与分类[J]. 图学学报, 2020, 41(3): 390-398.
[8]	陈国军，程琰，曹岳，李胜 . 基于目标特征的植株深度图像修复[J]. 图学学报, 2019, 40(3): 460-465.
[9]	夏永泉，王兵，支俊,黄海鹏，孙静茹. 基于随机森林方法的小麦叶片病害识别研究[J]. 图学学报, 2018, 39(1): 57-62.
[10]	李大华，赵相飞，许亮，于波. 基于纹理特征与 HSI 空间的苹果识别与标定[J]. 图学学报, 2016, 37(5): 688-693.
[11]	余春艳，苏晨涵. 视频语义上下文标签树及其结构化分析[J]. 图学学报, 2015, 36(5): 747-755.
[12]	杜柳青，余永维. 磁瓦表面缺陷机器视觉检测与识别方法[J]. 图学学报, 2014, 35(4): 590-594.
[13]	王向东，张静文，毋立芳，徐文泉. 一种运动轨迹引导下的举重视频关键姿态提取方法[J]. 图学学报, 2014, 35(2): 256-261.
[14]	伍尤富. 一种基于核主成分分析和组合分类器的虹膜识别方法[J]. 图学学报, 2012, 33(3): 102-106.