基于KNN Matting算法的冷金属过渡焊接熔滴尺寸检测  

Droplet Size Detection in Cold Metal Transition Welding Based on KNN Matting Algorithm

作　　者：管森 邢彦锋 曹菊勇 杨夫勇 张小兵 Guan Sen;Xing Yanfeng;Cao Juyong;Yang Fuyong;Zhang Xiaobing(School of Mechanical and Automobile Engineering,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学机械与汽车工程学院,上海市201620

出　　处：《农业装备与车辆工程》2022年第9期59-62,共4页Agricultural Equipment & Vehicle Engineering

基　　金：上海市科委项目“高强度钢-铝合金CMT焊缝热塑性变形及偏差传递机理研究”(20ZR1422600)。

摘　　要：针对冷金属过渡焊接熔滴轮廓表征不明显、无法预测熔滴过渡行为的情况,提出一种基于KNN Matting的目标提取算法,并引入IG算法生成三区标志图(Trimap)作为预先步骤,加速了KNN Matting算法迭代进程,加快图像分割流程;生成最终的前景蒙版(Alpha Matte)用来提取熔滴图像轮廓。结果表明,此方法在冷金属过渡焊的开始与过渡过程熔滴轮廓拟合准确,Trimap的生成质量是影响熔滴图像轮廓提取的重要因素,KNN Matting算法中的K值根据生成的Trimap决定。In view of the situation that the droplet profile of existing cold metal transition welding is not obvious and the droplet transition behavior cannot be predicted,this paper proposes a target extraction algorithm based on KNN Matting,and introduces IG algorithm to generate Trimap as a pre-step,which accelerates the iterative process of KNN Matting algorithm and accelerates the image segmentation process.The final Alpha Matte was generated to extract the droplet image contour.The results show that this method is accurate in the beginning and transition process of cold metal transition welding,and the generation quality of Trimap is an important factor affecting the extraction of droplet image contour.The K value in KNN Matting algorithm is determined according to the generated Trimap.

关 键 词：熔滴过渡 KNN Matting算法 IG算法 三区标志图 前景蒙版 

分 类 号：TG401[金属学及工艺—焊接]