基于改进AKAZE算法的数控铣削刀具三维图像重建  

3D Image Reconstruction of CNC Milling Tool Based on Improved AKAZE Algorithm

作　　者：陈渊 王海雄[1] 易怀安 CHEN Yuan;WANG Haixiong;YI Huaian(Key Laboratory of Advanced Manufacturing and Automation Technology,School of Mechanical and Control Engineering,Guilin University of Technology,Guilin Guangxi 541006,China)

机构地区：[1]桂林理工大学机械与控制工程学院,广西高校先进制造与自动化技术重点实验室,广西桂林541006

出　　处：《机床与液压》2025年第4期68-72,共5页Machine Tool & Hydraulics

基　　金：国家自然科学基金地区科学基金项目(52065016)。

摘　　要：为了对铣削加工刀具磨损进行精确在线检测,使用改进AKAZE算法对数控铣刀切削部位的特征进行提取与匹配,从而实现铣刀三维图像的重建。搭建铣刀三维图像重建系统,实现了图像的稀疏点云重建、密集点云重建、三角网格模型重建和纹理映射功能;分析特征匹配算法的主要思想和实现步骤,最后通过实验验证所提方法的可行性。结果表明:改进的CAKE算法提高了铣刀三维重建的效率,重建时间减少到1.15 s;与SIFT、SURF、AKZE算法相比,重建过程中稀疏点云的数量增加了8.3%、2%、13.7%,密集点云数量增加了11.8%、6.4%、17.7%,三角网格数增加了2.7%、1.7%和3.7%,模型质量相比其他算法明显提高。In order to accurately online monitor the milling tool wear,the improved AKAZE(accelerated-KAZE)algorithm was used to extract and match the features of the cutting part of the CNC milling cutter,so as to realize 3D image reconstruction for the milling cutter.The 3D image reconstruction system of milling cutter was constructed,and the sparse point cloud reconstruction,dense point cloud reconstruction,triangular mesh model reconstruction and texture mapping functions of the image were realized.Then,the main idea of the feature matching algorithm and the implementation steps were analyzed,and finally the feasibility of the proposed method was verified through experiments.The results show that the improved AKAZE algorithm improves the efficiency of 3D reconstruction of milling cutter,and the reconstruction time is reduced to 1.15 s;compared with the SIFT,SURF and AKAZE algorithm,the number of sparse point clouds in the reconstruction process is increased by 8.3%,2%and 13.7%,the number of dense point clouds is increased by 11.8%,6.4%and 17.7%,the number of triangular mesh is increased by 2.7%,1.7%and 3.7%,resulting in a significant improvement in model quality compared to other algorithms.

关 键 词：三维图像 特征提取 三维重建 AKAZE算法 

分 类 号：TG714[金属学及工艺—刀具与模具]