Obstacle detection from IPM and super-homography

Abstract

Abstract-We present in this article a simple method to estimate an IPM view from an embedded camera. The method is based on the tracking of the road markers assuming that the road is locally planar. Our aim is the development of a freespace estimator which can be implemented in an Autonomous Guided Vehicle to allow a safe path planning. Opposite to most of the obstacle detection methods which make assumptions on the shape or height of the obstacles, all the scene elements above the road plane (particularly kerbs and poles) have to be detected as obstacles. Combined with the IPM tranformation, this obstacle detection stage can be viewed as the first stage of a free-space estimator dedicated to AGV in the complex urban environments.

Fast Obstacle Detection for Monocular Autonomous Mobile Robots

“…The approach of Zhou and Li [15] uses normalized homography calculated by tracking feature points for ground plane detection. Simond and Parent [16] use IPM and super-homography for obstacle detection. They employed edge based feature and perspective effects for distinguishing obstacles from the ground plane.…”

“…IPM is a classical and reliable approach for detecting obstacles in front of autonomous robots [12]- [14], [16]. The robot acquires images C t 1 and C t 2 at different times t 1 and t 2 .…”

A Visual Navigation Strategy Based on Inverse Perspective Transformation

“…The distorted zones of the predicted image corresponded to objects. Simond combined the IPM with the computation of the ground plane super-homography from road lines to discriminate obstacles from road in an autonomous guided vehicle application (Simond & Parent, 2007).…”

Robust ground plane detection for obstacle avoidance of mobile robots using a monocular camera

“…Due to the difficulty in generalized obstacle detection, many researchers propose to resolve the problem of floor or ground detection based on planar property of road or indoor environment [7][8][9][10][11][12]. Liang et al [7] proposed to use reciprocal-polar (RP) image rectification and ground plane segmentation by sinusoidal model fitting in RP-space to segment the ground plane from a mobile robot's visual field of view.…”

“…For pixel level obstacles representation, warping or inverse perspective transform (IPT) technique has been widely used in both stereo-based and monocular-based approaches [10][11][12][13][14]. This technique uses ground plane homography to warp one of the images.…”