桂林电子科技大学学报2024，Vol.44Issue(6)：568-578,11.DOI:10.16725/j.1673-808X.202260
基于深度学习的高精度相机相对位姿估计网络
High precision camera relative pose estimation network based on deep learning
摘要
Abstract
Relative pose estimation of camera is to calculate the relative position and pose of the camera during imaging.It is a key problem in the field of computer vision,such as image mosaic,3D reconstruction,SLAM and so on.In order to obtain the most ac-curate results,the traditional algorithm needs repeated iterative,so it has a large amount of calculation and time consuming.Most of the existing deep learning algorithms take the left and right images as the input,and obtain the pose parameters based on the seman-tic features of pixels,so it has a large amount of data and complex model structure.To solve the above problems,a new deep learn-ing network for camera relative pose estimation was proposed,which took the correspondences as the input of network.After obtain-ing the correspondences between two images,firstly,the correspondences were divided into inliers(the correspondences with small matching error)and outliers(the others correspondences with large matching error)by a classification network.Then,taking all in-liers as inputs,the relative rotation and translation parameters of the camera were obtained quickly by a calculation network of cam-era relative pose parameter.The experimental results show that the proposed method is 1.9 times faster than the traditional algo-rithm and has less error;The new algorithm has better accuracy precision than the existing deep learning algorithm based on seman-tic features of pixels,but it processes less data and has a lighter network structure;At the same time,the designed network structure can adapt to the input of different number of correspondences.关键词
相对位姿估计/深度学习/自注意力机制/残差网络/最大池化Key words
relative pose estimation/deep learning/self-attention mechanism/residual network/max pooling分类
信息技术与安全科学引用本文复制引用
彭智勇,吴磊,肖博..基于深度学习的高精度相机相对位姿估计网络[J].桂林电子科技大学学报,2024,44(6):568-578,11.基金项目
广西自然科学基金(2020GXNSFAA159091) （2020GXNSFAA159091）
桂林电子科技大学研究生教育创新计划(2022YCXS157) （2022YCXS157）
