Incremental disparity space image computation for automotive applications

Abstract

Abstract-Generating a depth map from a pair of stereo images is a challenging task, which is often further complicated by the additional restrictions imposed by the target application; in the automotive field, for example, real-time environment reconstruction is essential for safety and autonomous navigation systems, thus requiring reduced processing times, often at the expense of a somewhat limited degree of accuracy in the results. Nevertheless, a-priori knowledge on the intended use of the algorithm can also be exploited to improve its performance, both in terms of precision and computational burden.This paper presents three different approaches to incremental Disparity Space Image (DSI) computation, which leverage the properties of a stereo-vision system installed on a vehicle to produce accurate depth maps at sustained frame rates on commodity hardware.

Stereo obstacle detection in challenging environments: The VIAC experience

“…Given the tight development schedule, Disparity Space Image (DSI) generation was implemented exploiting the already available window-based SAD correlation technique described in [9]. As reported in Sec.…”

Intelligent Overhead Sensor for Sliding Doors: A Stereo Based Method for Augmented Efficiency

“…The two background images are compared against input images performing a difference: binarized images (masks) are then generated after thresholding and dilating the difference images. A Sobel filter is used to reduce the changing light conditions problem and to produce an edges map that is used to generate Disparity Space Image (DSI) [3]. Thanks to the preliminary use of the IPM, the resulting DSI has to be interpreted using a specific approach (see Section 3)…”

“…5.b). In fact the algorithm for the implementation of DSI [3] requires to search for each pixel of the right image the corresponding pixel on the left image (homologous pixel). In IPM images, the homologous pixels are searched for in the same row in each images.…”

3DV — An embedded, dense stereovision-based depth mapping system

“…Stereovision-based depth mapping is a widely studied subject, which has seen substantial research [1], [2], [3], [4], [5] and benchmarking [6], [7], [8] efforts. The various algorithms offer different trade-offs in terms of computational complexity, reconstruction quality and robustness against noise in the input images.…”