盖革APD激光雷达高空间分辨率三维重构方法研究

Research on 3D reconstruction method for Geiger-mode APD laser radar with high spatial resolution

盖革模式阵列APD激光雷达由于像元数的限制导致重构的三维距离像分辨率低,严重限制了对场景信息的感知识别能力。为此,提出一种基于改进图像引导的盖革模式阵列APD激光雷达超分辨三维图像重构算法。主要研究基于配准后的盖革APD低分辨距离图像、增强电荷耦合器件(ICCD)相机高分辨强度图像和马尔科夫随机场模型,建立具有距离保真项和正则化项约束的全局能量函数,最终对优化模型的最优化求解实现高分辨强度图像对低分辨距离图像的超分辨引导重构。实验结果表明,相较于现有算法,文中算法在提升重构三维图像分辨率的同时,也使图像的边缘更加锐利。该研究对于盖革模式阵列APD激光雷达的目标探测、识别和跟踪具有重要意义。

Due to the limitation of pixel number of Geiger-mode avalanche photodiode(Gm-APD)array laser radar,the resolution of reconstructed 3D range image is low,which seriously limits its perception and recognition ability of scene information.In view of the above,a Gm-APD array laser radar superresolution 3D image reconstruction algorithm based on improved image guidance is proposed.The research is mainly based on the registered Gm-APD low-resolution range image,highresolution intensity image by intensified charge-coupled device(ICCD)camera and Markov random field(MRF)model.The global energy function with the range fidelity term and regularization term constraints is established.Finally,the optimal solution of the optimized model is carried out to achieve the super-resolution guided reconstruction of high-resolution intensity image to low-resolution range image.The experimental results show that the proposed algorithm can improve the resolution of the reconstructed 3D image,and ensure sharper edges of the image in comparison with the existing algorithms.This research is of great significance for object detection,recognition and tracking of Gm-APD array laser radar.