随机蕨丛算法下的三点焊缝定位方法  

Three-Point Welding Seam Location Method Based on Random Ferns Algorithm

作　　者：高恩阳[1] 高齐竹慧 刘剑[1] 刘美菊[1] GAO Enyang;GAO Qizhuhui;LIU Jian;LIU Meiju(College of Information and Control Engineering, Shenyang Jianzhu University, Shenyang 110168, China)

机构地区：[1]沈阳建筑大学信息与控制工程学院,沈阳110168

出　　处：《电子器件》2018年第3期560-564,共5页Chinese Journal of Electron Devices

基　　金：国家自然科学青年基金项目(61305125)

摘　　要：针对在焊接过程中工件易变形、光照干扰等情况下不能精确定位焊缝的问题,提出一种基于方向梯度特征的随机蕨丛算法。首先建立焊缝模型分类器,并用其在测试样本中得到焊缝部分。其次,采用随机蕨丛算法得到所有训练样本的特征类型并计算特征点的平均坐标差值。最后,判断测试样本所属类型,根据坐标偏移量求出特征点具体坐标值来实现定位。实验结果表明:所提出的算法在保证较高精确定位的同时,还具有很好的时效性。Aiming at the problem that is not easy to accurately locate the weld seam when the workpiece is deformed or the situation of light interference,an algorithm based on histogram of oriented gridients is proposed for Random ferns. Firstly,the weld model classifier is established and the classifier is used to get the weld part in test samples.Secondly,the Random ferns is used to get the feature types of all the training samples,the feature points＇ average coordinate difference is calculated. Finally,we can determine the type of test samples,get the specific coordinate value to achieve the positioning according to the coordinate offset. The experimental results show that our algorithm can ensure the high precise positioning and good timeliness.

关 键 词：随机蕨丛 焊缝定位 半朴素贝叶斯 方向梯度特征 支持向量机 

分 类 号：TG409[金属学及工艺—焊接] TP391.41[自动化与计算机技术—计算机应用技术]