GROUND PLANE DETECTION FOR AUTONOMOUS ROBOTS IN COMPLEX ENVIRONMENTS INCLINED WITH FLEXED FAR-FIELD TERRAINS

摘要

Ground plane detection is often used as one of thernimportant safety key operations to address some of thernissues associated with autonomous navigation in complexrnenvironments. Despite the strides on related detectionrnmethods developed for such navigation, detection ofrnground planes inclined with flexed far-fields to alleviaternrobot short-sightedness, with a guarantee on tarred andrncoarse terrains have received little attention. Finding arnsolution to these uncertainty problems is a challenge. Inrnthis paper, collective intelligence of the EmergentrnSituation Awareness (ESA) technology is proposed as arnsupportive strategy for autonomous robotic navigation.rnThe ability to reveal uncertainties over time on flexed farfieldrnis a ground plane detection strategy embedded in therncomplex environments. Experimental evaluations of thernESA by benchmarking the results of publicly availablernroads promise that collective intelligence will one day putrnan end to most autonomous ground plane detectionrnproblems. Such detection on flexed far fieldsrntremendously contributes to good navigational strategiesrnfor robotic vehicles being cautious of road accidents.