Navigation capability in complex and unknown outdoor environments is one of the major requirements for an autonomous vehicle and a robot that performs tasks such as a military mission or planetary exploration. Robust traversability estimation in unknown environments would allow the vehicle to devise control and planning strategies to maximize their effectiveness. In this study, we present a self-supervised on-line learning architecture to estimate the traversability in complex and unknown outdoor environments. The proposed approach builds a model by clustering appearance data using the newly proposed incremental nonparametric Bayesian clustering algorithm. The clusters are then classified as being either traversable or non-traversable. Because our approach effectively groups unknown regions with similar properties, while the vehicle is in motion without human intervention, the vehicle can be deployed to new environments by automatically adapting to changing environmental conditions. We demonstrate the performance of the proposed clustering algorithm through intensive experiments using synthetic and real data and evaluate the viability of the traversability estimation using corrected real data sets.
Related publications
1. H Lee, K Kwak, S Jo, An incremental nonparametric Bayesian clustering-based traversable region detection, Autonomous Robots 41(4), 795-810, 2017. [LINK][PDF]
2. H Lee, K Kwak, S Jo, Multisensor-based online positive learning for drivable region detection, Electronics Letters, 50(20), 2014. [LINK][PDF]
3. 이홍구, 조성호, 다변량 가우시안 클러스터링 기법을 이용한 비구조화 지면 추출, 한국정보과학회논문지:컴퓨팅의 실제 및 레터 (초청논문) , 제19권 4호, 2013년.
4. 이홍구, 조성호, 복셀 피처와 다변량 가우시안 클러스터링 기법을 이용한 비구조화 지형 모델링, 제 20회 지상무기학술대회 2012년12월.
5. 이홍구, 조성호, 다변량 가우시안 클러스터링을 이용한 지면 추출 기법,한국정보과학회, 제 39회 정기 총회 및 추계학술 발표회 2012년11월 (우수논문)
