Obstacle Detection Quality as a Problem-Oriented Approach to Stereo Vision Algorithms Estimation in Road Situation Analysis

Abstract

In this work we present a method for performance evaluation of stereo vision based obstacle detection techniques that takes into account the specifics of road situation analysis to minimize the effort required to prepare a test dataset. This approach has been designed to be implemented in systems such as self-driving cars or driver assistance and can also be used as problem-oriented quality criterion for evaluation of stereo vision algorithms.

L-shape fitting algorithm for 3D object detection in bird’s-eye-view in an autonomous driving system

“…Fusion can be divided into three categories [13]: high-level fusion, where measurements from each sensor are used independently to detect objects and the obtained detections are then fused; low-level fusion, which fuses raw measurements from sensors and use all of the collected data for detection; and midlevel fusion, where contextual features are extracted from raw measurements and then fused for further processing. An example of fusing (specifically, low-level fusing) is [14], where a depth map is constructed from two RGB images and then projected into three-dimensional coordinates.…”