3D point cloud-based place recognition is highly demanded by autonomous driving in GPS-challenged environments and serves as an essential component (i.e. loopclosure detection) in lidar-based SLAM systems. This paper proposes a novel approach, named NDT-Transformer, for realtime and large-scale place recognition using 3D point clouds. Specifically, a 3D Normal Distribution Transform (NDT) representation is employed to condense the raw, dense 3D point cloud as probabilistic distributions (NDT cells) to provide the geometrical shape description. Then a novel NDT-Transformer network learns a global descriptor from a set of 3D NDT cell representations. Benefiting from the NDT representation and NDT-Transformer network, the learned global descriptors are enriched with both geometrical and contextual information. Finally, descriptor retrieval is achieved using a query-database for place recognition. Compared to the state-of-the-art methods, the proposed approach achieves an improvement of 7.52% on average top 1 recall and 2.73% on average top 1% recall on the Oxford Robotcar benchmark. †Authors contribute equally.This project is funded by EPSRC FAIR-SPACE Hub (EP/R026092/1), Royal Society RGS/R2/202432 and EU Horizon 2020 ILIAD (No. 732737).

No abstract

Abstract. In the process of infrared thermal image segmentation of distribution equipment, efficiency of the segmentation is low; it is easy to appear the over-segmentation phenomenon and cannot better segment the anomaly region of infrared thermal. To solve this problem, starting from the basic principle of the watershed algorithm, an improved watershed algorithm based on morphological marker was proposed. Firstly, the algorithm operate morphological gradient for image to be segmented. Then it extracts the region minimum by using H-minima transform and use minima imposition technology to remark the gradient image. In the end, the marked gradient image was segmented by watershed algorithm. According to the experimental verification of MATLAB, this improved segmentation method has higher feasibility and can successfully extract anomaly region of the infrared thermal image.

No abstract