Pedestrian Protection Pedestrian Detection Vulnerable Road Users PROTECTOR SAVEU

Main long-term research theme at Daimler Research has been the visual detection of pedestrians from a moving vehicle. Here is a synopsis of my published work and other material from the public domain.

Pedestrians are arguably the most vulnerable traffic participants.

Article on the Daimler pedestrian system
(abstract, full text)

DaimlerChrysler Pedestrian Classification Benchmark

Survey article on pedestrian detection
(full text)

	Killed	Injured	Total
Passenger Cars	22.502	995.026	1.017.528
Pedestrians	6.049	155.151	161.200
Mopeds	2.421	141.870	144.291
Bicycles	2.385	139.442	141.827
Motor Cycles	3.821	124.023	127.844
Other	4.559	121.816	126.375
Total	41.737	1.677.328	1.719.065


	Road traffic accidents 1997. Total figures for EU countries Accident Source: UN-ECE)


	More than 150.000 pedestrians are injured yearly in the EU.

		More than 6000 are killed.




	Children are especially at risk.


		© Daimler

	The EU has recognized the importance of the problem and plans to phase-in legislation regarding the maximum tolerated leg- and head- impact coefficients in vehicle-pedestrian crashes at vehicle speeds of 40 km/h (phase 1 for 2005 already agreed with the automobile industry, phase 2 planned for 2010 still subject of debate). The automobile industry has already responded with a package of voluntary measures to improve pedestrian protection, i.e. automatic headlight activation, ABS/EPS systems and removal of rigid frames mounted on vehicle front.

Future pedestrian protection measures will likely include extensible vehicle structures (e.g. hood, bumper) which expand during collision in order to minimize impact of the pedestrian leg or head hitting the vehicle. One such prototype system was for example presented by Autoliv.

The aim is to develop vision-based driver assistance systems which detect dangerous situations involving pedestrians ahead of time, allowing the possibility to warn the driver or to brake the vehicle. Such systems are particularly valuable when the driver is distracted or visibility is poor.

Yet vision-based pedestrian detection is a difficult problem for a number of reasons. The objects of interest appear in highly cluttered backgrounds and have a wide range of appearances, due to body size and pose, clothing and outdoor lighting conditions. They stand typically far away from the camera, and thus appear rather small in the image, at low resolution. A major complication is that because of the moving vehicle, one does not have the luxury to use simple background subtraction methods (such as those used in surveillance applications) to obtain a foreground region containing the human. Finally, there are hard real-time requirements and stringent performance criteria.

See right for a diagram of the Daimler pedestrian system. The detection component consists of a cascade of module, each utilizing different visual criteria to successively focus on relevant image regions, carefully balancing robustness and efficiency considerations. The tracking component aggregates per-frame detections to trajectories by a tracking module. Finally, the risk assessment and warning/control component evaluates the probability of collision; if the latter exceeds a threshold an acoustic driver warning is given or automatic vehicle braking is applied.

Pedestrians are detected in a range 5-25m and up to 4m lateral, on each side of vehicle. It does so at processing rates of 7-15 Hz, allowing vehicle speeds up to 40 km/h.


	The system was integrated in a Mercedes-Benz E-Class limousine.


	Below an example of the system output: the left window shows the rectangular regions of interest as a result of stereo preprocessing. The middle window shows the final detection superimposed in red on the image. The right window shows a topic view of the situation in front of the vehicle. Right scale in meters. Red dots are current pedestrian locations, blue dots are pedestrian locations at previous time instants. Velocities are also drawn in (positions and velocities are both taken with respect to the moving vehicle).


			© Daimler

4,4 Mb (mpg)

Here is a video clip of the system in action on the test track and in real urban traffic. Besides the correct detections also a “false positive” is visible on the side of a car.

Research on vision-based pedestrian protection was previously conducted in EU 5th Framework projects PROTECTOR (2000-2003) and SAVE-U (2002-2005). It now continues under the EU 6th Framework WATCH-OVER (2005-2008) and the German AKTIV-SFR (2006-2010) projects. More details ...

PROTECTOR web site

Although obtained results are promising, more research is needed before such pedestrian systems can be placed at the hands of ordinary vehicle drivers. Yet it is another step closer towards Daimler’s long-term Vision of Accident-Free Driving.


	[Home Gavrila] [Resume] [Univ. Amsterdam] [Computer Vision] [Publications] [Media Coverage] [Search]


	Copyright © 2001-2006 Gavrila. All rights reserved. Disclaimer.