基于EAST的线激光条纹中心提取算法  

Algorithm of Line Laser Stripe Center Extraction Based on EAST

作　　者：贺锋涛 赵伟琳 He Fengtao;Zhao Weilin(College of Electronic Engineering,Xi'an University of Posts and Telecommunications,Xi'an 710121,Shaanxi,China)

机构地区：[1]西安邮电大学电子工程学院,陕西西安710121

出　　处：《应用激光》2025年第2期188-196,共9页Applied Laser

基　　金：装备预研教育部联合基金资助项目(8091B032130);西安邮电大学研究生联合培养工作基地(YJGJ201905)。

摘　　要：提出一种新的线结构光中心提取算法。针对传统分割算法对图像分割出现空洞、分割不连续以及不精确等问题,采用基于EAST深度学习模型对激光条纹的感兴趣区域进行准确高效提取,避免传统算法在分割过程中易受噪声干扰及效率低的不足。此外,采用灰度重心法对所检测的有效光条区域进行初提取,最后使用优化算法对初始中心线进行二次优化,实现光条中心亚像素坐标的精确提取。结果表明,该算法能够更加准确地提取出线结构光光条中心,保证条纹中心提取的精度与稳定性。This paper presents a new algorithm for center extraction of line structured light.Traditional segmentation algorithms often suffer from issues such as holes,discontinuity,and inaccuracy in image segmentation.To address these limitations,the EAST deep learning model is employed to accurately and efficiently extract the region of interest of laser stripes.This approach mitigates the susceptibility of traditional algorithms to noise interference and improves segmentation efficiency.In addition,the gray barycenter method is used to initially extract the detected effective light strip area,and finally the optimization algorithm is used to perform secondary optimization on the initial center line to realize the accurate extraction of the sub-pixel coordinates of the light strip center.The results show that the algorithm can more accurately extract the center of the line-structured light strip,which ensures the accuracy and stability of the strip center extraction.

关 键 词：三维重建 深度学习 中心提取 曲线拟合 优化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]