Paolo Zani

The Single Frame Stereo Vision System for Reliable Obstacle Detection Used during the 2005 DARPA Grand Challenge on TerraMax

Autonomous driving in off-road environments requires an exceptionally capable sensor system, especially given that the unstructured environment does not provide many of the cues available in on-road environments. This paper presents a variable-width-baseline (up to 1.5 meters) single-frame stereo vision system for obstacle detection that can meet the needs of autonomous navigation in extreme environments. Efforts to maximize computational speed oth in the attention given to accurate and stable calibration and the exploitation of the processors MMX and SSE instruction sets - allow a guaranteed 15 fps rate. Along with the assured speed, the system proves very robust against false positives. The system has been field tested on the TerraMax vehicle, one of only five vehicles to complete the 2005 DARPA Grand Challenge course and the only one to do so using a vision system for obstacle detection

Autonomous vehicles control in the VisLab Intercontinental Autonomous Challenge

Autonomous driving is one of the most interesting fields of research, with a number of important applications, like agricultural, military and, most significantly, safety. This paper addresses the problem of designing a general purpose path planner and its associated low level control for autonomous vehicles operating in unknown environments. Different kinds of inputs, like the results of obstacle detection, ditch localization, lane detection, and global path planning information are merged together using potential fields to build a representation of the environment in real-time; kinematically feasible trajectories, based on vehicle dynamics, are generated on a cost map. This approach demonstrated both flexibility and reliability for vehicle driving in very different environments, including extreme road conditions. This controller was extensively tested during VIAC, the VisLab Intercontinental Autonomous Challenge, a 13,000 km long test for intelligent vehicle applications. The results, collected during the development stage and the experiment itself, are presented in the final part of this article.

TerraMax Vision at the Urban Challenge 2007

This paper presents the TerraMax vision systems used during the 2007 DARPA Urban Challenge. First, a description of the different vision systems is provided, focusing on their hardware configuration, calibration method, and tasks. Then, each component is described in detail, focusing on the algorithms and sensor fusion opportunities: obstacle detection, road marking detection, and vehicle detection. The conclusions summarize the lesson learned from the developing of the passive sensing suite and its successful fielding in the Urban Challenge.

VIAC: An out of ordinary experiment

This paper presents the preliminary results of VIAC, the VisLab Intercontinental Autonomous Challenge, a test of autonomous driving along an unknown route from Italy to China. It took 3 months to run the entire test; all data have been logged, including all data generated by the sensors, vehicle data, and GPS info. This huge amount of information has been packed during the trip, compressed, and transferred back to Parma for further processing. This data is now ready for a deep analysis of the various systems performance, with the aim of virtually running the whole trip multiple times with improved versions of the software. This paper discusses some preliminary figures obtained by the analysis of the data collected during the test. More information will be generated by a deeper analysis, which will take additional time, being the data about 40 terabyte in size.

GOLD: A Framework for Developing Intelligent-Vehicle Vision Applications

To develop real-time vision applications for use in highly dynamic environments, such as automotive traffic, researchers must gather large amounts of data from different sensors and systems at different rates. Software capable of real-time data acquisition, synchronization, logging, and - ultimately - data processing and visualization is fundamentally important to improving researcher efficiency. The general obstacle and lane detection framework supports different devices and makes it easy to add new system functionalities. GOLD can easily become the engine for many automotive applications, and it could work in other application domains as well.

Robust monocular lane detection in urban environments

An effective lane detection algorithm is a basic, yet fundamental component of both autonomous navigation and advanced road safety systems; this paper presents an approach that produces reliable results exploiting a robust polyline matching technique. The proposed solution has been designed from the ground up so that only very limited hardware resources are required: just one camera is used, and the processing is fast enough to be compatible with mainstream DSP units.

A full-3D voxel-based dynamic obstacle detection for urban scenario using stereo vision

Autonomous Ground Vehicles designed for dynamic environments require a reliable perception of the real world, in terms of obstacle presence, position and speed. In this paper we present a flexible technique to build, in real time, a dense voxel-based map from a 3D point cloud, able to: (1) discriminate between stationary and moving obstacles; (2) provide an approximation of the detected obstacles absolute speed using the information of the vehicles egomotion computed through a visual odometry approach. The point cloud is first sampled into a full 3D map based on voxels to preserve the tridimensional information; egomotion information allows computational efficiency in voxels creation; then voxels are processed using a flood fill approach to segment them into a clusters structure; finally, with the egomotion information, the obtained clusters are labeled as stationary or moving obstacles, and an estimation of their speed is provided. The algorithm runs in real time; it has been tested on one of VisLabs AGVs using a modified SGM-based stereo system as 3D data source.

Stereo obstacle detection in challenging environments: The VIAC experience

Obstacle detection by means of stereo-vision is a fundamental task in computer vision, which has spurred a lot of research over the years, especially in the field of vehicular robotics. The information provided by this class of algorithms is used both in driving assistance systems and in autonomous vehicles, so the quality of the results and the processing times become critical, as detection failures or delays can have serious consequences. The obstacle detection system presented in this paper has been extensively tested during VIAC, the VisLab Intercontinental Autonomous Challenge [1], [2], which has offered a unique chance to face a number of different scenarios along the roads of two continents, in a variety of conditions; data collected during the expedition has also become a reference benchmark for further algorithm improvements.

The passive sensing suite of the TerraMax autonomous vehicle

This paper presents the TerraMax autonomous vehicle, which competed in the DARPA Urban Challenge 2007. The sensing system is mainly based on passive sensors, in particular four vision subsystems are used to cover a 360deg area around the vehicle, and to cope with the problems related to complex traffic scenes navigation. A trinocular system derived from the one used during the 2005 Grand Challenge performs obstacle and lane detection, twin stereo systems (one in the front and one in the back) monitor the area close to the truck, two lateral cameras detect oncoming vehicles at intersections, and a rear view system monitors the lanes next to the truck looking for overtaking vehicles. Data fusion between laser scanners and vision will be discussed, focusing on the benefits of this approach.

An evaluation of monocular image stabilization algorithms for automotive applications

The performance of many computer vision applications, especially in the automotive field, can be greatly increased if camera oscillations induced by movements of the acquisition devices are corrected by a stabilization system. An effective stabilizer should cope with different oscillation frequencies and amplitude intervals, and work in a wide range of environments (such as urban, extra-urban or even unstructured ones). In this paper we analyze three different approaches, based on signature, feature, and correlation tracking respectively, that have been devised to face these problems