Perception for collision avoidance and autonomous driving

Abstract

The Navlab group at Carnegie Mellon University has a long history of development of automated vehicles and intelligent systems for driver assistance. The earlier work of the group concentrated on road following, cross-country driving, and obstacle detection. The new focus is on short-range sensing, to look all around the vehicle for safe driving. The current system uses video sensing, laser rangefinders, a novel light-stripe rangefinder, software to process each sensor individually, a map-based fusion system, and a probability based predictive model. The complete system has been demonstrated on the Navlab 11 vehicle for monitoring the environment of a vehicle driving through a cluttered urban environment, detecting and tracking fixed objects, moving objects, pedestrians, curbs, and roads. Robotics Technology and Intelligent VehiclesDuring the decade of the 1990's, researchers from automotive engineering, from robotics, and from allied fields came together to start building intelligent vehicles [1,2,3,4]. These vehicles use sensors, actuators, processors, and communications systems to either drive automatically, or to monitor a human driver and assist in navigation or warn in case of a developing dangerous situation.The field was driven by several market demand factors. First is an increasing interest in safety. While the accident rate has decreased steadily for 40 years, the total miles traveled have increased, with the net result that the number of road fatalities in the US has stayed approximately constant, averaging 40,000 / year [5]. Furthermore, there is an increasing sense that vehicles have become safer (air bags, anti-lock brakes, crumple zones); roadways have become safer (the Interstate highway system); societal pressures have become stronger (mandatory child safety seats, serious enforcement of drunk driving laws) but the drivers themselves have not necessarily improved. Currently, in the US, 70% of all crashes are primarily caused by the drivers, and in another 20% driver error is a major contributing factor [5,6].Beyond safety, the next major driving factors have to do with congestion and pollution. In the US, traffic continues to grow at 4% / year, while with the completion of the Interstate highway system, major new road construction greatly lags traffic growth. Computercontrolled vehicles promise to smooth out traffic flow, permitting more vehicles to travel in

Self-supervised Monocular Road Detection in Desert Terrain

“…In addition to using tracking of lane markings, CMU also deals with inner-city driving involving cluttered scenes and collision avoidance [9].…”

Real-time coordination of autonomous vehicles

“…Research in this area includes the search for adequate sensors and actuators [6], vehicle control algorithms [4], and assessment of the usability of such vehicles [2]. This has led to a number of results, including a practical demonstration of driverless vehicles following a road lane, overtaking a slower vehicle, and crossing an unsignalised junction [3], [7].…”

Vision-based Motion Planning for an Autonomous Motorcycle on Ill-Structured Road

“…If a robot is running on a well-structured road, such as freeways or the roads in an urban area [3], the primary focus of research is lane detection [4] using surface and boundary features, and road following [5], which detects road trends. Since the road has a relatively uniform surface and clear lane markings, techniques such as road segmentation, road edge detection [6], and curve-fitting [7] are often used to generate vehicle control inputs.…”

“…Motion blurring and vibration caused by a fast moving vehicle further degrade image quality. To address these issues, researchers approach the problem using different strategies such as color vision [10], [16], prior knowledge [6], pixel voting [15], classifier fusion [14], optical flow [21], neural networks [3], and machine learning [20], [21].…”