基于Voronoi算法的卸船机抓斗骨架机器视觉识别  

Machine Vision Recognition of Grab Bucket Skeleton of Ship Unloader Based on Voronoi Algorithm

作　　者：杨红民 Yang Hongmin(China Shenhua Energy Co.,Ltd.,Beijing 100011,China)

机构地区：[1]中国神华能源股份有限公司,北京100011

出　　处：《港口航道与近海工程》2024年第3期17-21,共5页Port,Waterway and Offshore Engineering

摘　　要：近年来,机器视觉识别依旧是炙手可热的话题,骨架是基于物体形状特征的简化对象描述方式,得到理想的物体骨架具有重要的意义。卸船机的抓斗骨架视觉识别是抓斗卸船机全自动能否完成预定目标的重要影响因素,机器视觉识别并非是在处理线性的数据,即识别对象和识别背景是随着时间的变化而变化的。骨架作为物体形状的一种表示方式,与采用轮廓描述形状相比,拥有许多优点。该方法将目标的外形特征与最大圆弧、主干等相关,便于目标形态的保持与重建。总的来说,基于Voronoi算法的卸船机抓斗骨架识别成功率有了较大的提升。In recent years,machine vision recognition is still kept a hot topic.and skeleton is a simplified description method based on the features of object shape.Obtaining an ideal object skeleton is of great significance.The visual recognition of the grab bucket skeleton of ship unloader is an important factor for fully automatic grab bucket unloader achieving the preset goal.Machine vision recognition applies to non-linear data,that is,the recognition object and background change over time.As a representation of the shape of an object,skeleton has many advantages in comparison with contour description method.This method associates the shape features of the object with the maximum arc,trunk,etc.,which facilitates the preservation and reconstruction of the target shape.Overall,the success rate of identifying the grab bucket skeleton of ship unloader has been significantly improved based on Voronoi algorithm.

关 键 词：机器视觉 抓斗骨架识别 Voronoi算法 桥式全自动卸船机 轮廓特征 轮廓起始点配准 

分 类 号：TP273.5[自动化与计算机技术—检测技术与自动化装置]