基于方向感知网络的场景偏振三维成像技术(特邀)  

Scene Polarization 3D Imaging Technology Based on Direction-Aware Network(Invited)

作　　者：张俊杰[1] 刘飞[1] 王鑫[2] 何飘 樊昭 邵晓鹏[1,2] Zhang Junjie;Liu Fei;Wang Xin;He Piao;Fan Zhao;Shao Xiaopeng(School of Optoelectronic Engineering,Xidian University,Xi’an 710071,Shaanxi,China;Hangzhou Institute of Technology,Xidian University,Hangzhou 311231,Zhejiang,China)

机构地区：[1]西安电子科技大学光电工程学院,陕西西安710071 [2]西安电子科技大学杭州研究院,浙江杭州311231

出　　处：《激光与光电子学进展》2024年第2期415-424,共10页Laser & Optoelectronics Progress

基　　金：国家自然科学基金(62205259,62075175,61975254,62375212,62005203,62105254);西安电子科技大学基本科研业务费资助项目(QTZX22016)。

摘　　要：针对场景偏振三维成像中光照不均匀、色彩、材料复杂和大视场下观测方向变化等原因造成的偏振法线梯度不准确和真实三维信息获取困难的问题,提出一种基于方向感知卷积神经网络的场景偏振三维成像新方法。首先,搭建具有方向感知能力的场景深度估计网络结构;其次,利用卷积神经网络所估计的场景深度对偏振法线梯度进行校正;最后,利用校正后的梯度通过基于梯度的积分算法进行三维重建。实验结果表明,所提方法解决了偏振固有的方位角模糊,提高了在光照不均匀、大视场范围场景条件下获取的法线梯度的准确性,最终在恢复场景真实三维形状的同时保留了丰富的纹理细节信息。实验结果证明了所提技术的有效性与优越性。To overcome challenges arising from inaccurate polarization normal gradients and difficulties in obtaining real three-dimensional(3D) information in scene polarization 3D imaging—attributed to factors like uneven illumination,complex colors,materials,and changes in observation direction under a large field of view—a new approach utilizing a direction-aware convolution neural network is explored.The method involves constructing a scene depth estimation network with direction perception abilities,correcting polarization normal gradients using the convolutional neural network estimated scene depth,and ultimately reconstructing the 3D image through a gradient-based integration algorithm.Experimental results showcase this approach's effectiveness in resolving azimuth ambiguity inherent in polarization,enhancing normal gradient accuracy in scenes with uneven illumination and wide field of view,and successfully restoring the real 3D shape of the scene while preserving intricate texture details.The findings affirm the efficacy and superiority of the proposed technology.

关 键 词：偏振三维成像 深度估计 梯度场校正 神经网络 

分 类 号：O436.3[机械工程—光学工程]