Novel Perception Algorithmic Framework For Object Identification and Tracking In Autonomous Navigation

TL;DR

This paper presents a training-free perception framework for object identification and tracking in autonomous vehicles, utilizing geometric and probabilistic methods to achieve high accuracy and real-time performance.

Contribution

It introduces a novel perception algorithm that combines pose estimation, KD-Tree segmentation, and Bayesian tracking without requiring training data.

Findings

Median tracking accuracy of 91% on KITTI dataset

End-to-end processing time of 153 milliseconds

Effective in real-time autonomous navigation scenarios

Abstract

This paper introduces a novel perception framework that has the ability to identify and track objects in autonomous vehicle's field of view. The proposed algorithms don't require any training for achieving this goal. The framework makes use of ego-vehicle's pose estimation and a KD-Tree-based segmentation algorithm to generate object clusters. In turn, using a VFH technique, the geometry of each identified object cluster is translated into a multi-modal PDF and a motion model is initiated with every new object cluster for the purpose of robust spatio-temporal tracking. The methodology further uses statistical properties of high-dimensional probability density functions and Bayesian motion model estimates to identify and track objects from frame to frame. The effectiveness of the methodology is tested on a KITTI dataset. The results show that the median tracking accuracy is around 91% with an end-to-end computational time of 153 milliseconds