Learning terrain segmentation with classifier ensembles for autonomous robot navigation in unstructured environments

Abstract

Autonomous robot navigation in unstructured outdoor environments is a challenging area of active research and is currently unsolved. The navigation task requires identifying safe, traversable paths that allow the robot to progress toward a goal while avoiding obstacles. Stereo is an effective tool in the near field, but used alone leads to a common failure mode in autonomous navigation in which suboptimal trajectories are followed due to nearsightedness, or the robot's inability to distinguish obstacles and safe terrain in the far field. This can be addressed through the use of machine learning methods to accomplish near-to-far learning, in which near-field terrain appearance and stereo readings are used to train models able to predict far-field terrain. This paper proposes to enhance existing, memoryless near-to-far learning approaches through the use of classifier ensembles that allow terrain models trained on data seen at different points in time to be preserved and referenced later. These stored models serve as memory, and we show that they can be leveraged for more effective far-field terrain classification on future images seen by the robot. A five-factor, full-factorial, repeated-measures experimental evaluation is performed on hand-labeled data sets taken directly from the problem domain. The experiments result in many statistically significant findings, the most important being that the proposed near-to-far Best-K Ensemble Algorithm, with appropriate parameter selection, outperforms the single-model, nonensemble baseline approach in far-field terrain classification. Several other findings that inform the use of near-to-far ensemble methods are also presented.

Obstacle Classification and 3D Measurement in Unstructured Environments Based on ToF Cameras

“…However, it can provide high-resolution images, texture and color information, not just a 3D geometry. Now, stereo cameras have been widely applied to map-building [8–12], natural environment navigation, and rescue missions [13]. Nevertheless, stereo vision has limitations, especially in environments with few features and limited visibility due to undesirable light conditions.…”

Adaptive cascade of boosted ensembles for face detection in concept drift

“…The incoming data are high-speed and large scale, with drifting concepts caused by different lighting conditions. A problem domain of this kind requires real-time processing and adaptation in order for time-sensitive tasks such as obstacle avoidance and planning to be effective [12].…”

Towards a fully-autonomous vision-based vehicle navigation system in outdoor environments

“…Standard classification algorithms yield poor results when the training data is unbalanced [8]. In order to solve the problem of unbalanced data, Procopio et al [9] randomly select a predetermined number of samples for both traversable and non-traversable classes. This method eliminates elements from majority class to match the size of the minority class and if the number of available samples in either class is less than the target number of training examples, learning is not performed for that frame and the vehicle navigates only based on its near-field stereo information.…”