Fast Object Classification and Meaningful Data Representation of Segmented Lidar Instances

TL;DR

This paper presents a computationally efficient method for real-time Lidar object classification on CPU, using segmented instances and a novel data representation to enable deployment in embedded vehicle systems.

Contribution

The authors introduce a new data representation and classification approach that allows real-time Lidar object classification on CPU without specialized hardware.

Findings

Capable of real-time processing on CPU

Achieves good results on public datasets

Applicable to object detection and segmentation tasks

Abstract

Object detection algorithms for Lidar data have seen numerous publications in recent years, reporting good results on dataset benchmarks oriented towards automotive requirements. Nevertheless, many of these are not deployable to embedded vehicle systems, as they require immense computational power to be executed close to real time. In this work, we propose a way to facilitate real-time Lidar object classification on CPU. We show how our approach uses segmented object instances to extract important features, enabling a computationally efficient batch-wise classification. For this, we introduce a data representation which translates three-dimensional information into small image patches, using decomposed normal vector images. We couple this with dedicated object statistics to handle edge cases. We apply our method on the tasks of object detection and semantic segmentation, as well as the relatively new challenge of panoptic segmentation. Through evaluation, we show, that our algorithm is capable of producing good results on public data, while running in real time on CPU without using specific optimisation.