A.3.0 Road vehicle guidance (by ‘dynamic vision’ at UniBwM / LRT / ISF)



A.3.1 Road recognition and vehicle guidance

·        Shape description by curvature (change of direction over arc length)

·        recognition by aggregated edge features (smooth continuation)

·        relative ego-state from features nearby (in lower part of image)

·        expectation-guided, efficient feature extraction including steer angle control output.

Vehicle guidance by feed-forward control from curvature and super-imposed state feedback with integral component of lateral offset.

 



A.3.2 Skidpan demo to Daimler-Benz AG 1986

First demonstration of dynamic vision capabilities for road vehicle guidance to Daimler-Benz Research:

·        Longitudinal & lateral guidance with lateral acceleration feedback;

  • speed range from 0 to 36 km/h (10 m/s), autonomous reduction to Vmax,curve  = 7 m/s circling speed with radius ~ 50 m, yielding ~ 0.1 g lateral acceleration.

^ click on image opens high quality picture ^

A.3.3 Various types of roads with / without lane markings:

·        High-speed roads with multiple lanes and lane markings,

·        sealed roads without lane markings, with & without curbstones,

·        shadows from trees and buildings in all seasons.

  • Unsealed roads with noisy transition to shoulders.




^ click on image opens high quality picture ^

A.3.4 Lane change on a free road since 1987, situation-dependent feed-forward control time history with super-imposed (delayed) feedback in new lane. Autonomous lane change decision 1994.

 

^ click on image opens high quality picture ^

 

A.3.5 Perception of horizontal and vertical curvature decoupled by differential geometry formulation (vertical profile above the horizontal one), ~ constant road width for observability;

 



A.3.6:

A.3.6.1 to A.3.6.7 Obstacle detection and avoidance



A.3.6.1-2a Single obstacle detection and avoidance

A.3.6.1 VaMoRs with 1st-generation vision system BVV2

Monocular motion stereo, an integral part of the 4-D approach; Underlying assumption:

·     Gravity pulls both the obstacle and the own vehicle onto the flat ground;

·     view fixation by alternative centering of gaze direction horizontally and vertically on the vertical and horizontal centers of edge features.

·     Approach speed up to ~ 40 km/h.

A.3.6.2aStop & Go’ behind another road vehicle (Prometheus 1991)

A.3.6.2bc Multiple obstacle detection and avoidance

A.3.6.2b Cruise control with multiple vehicle detection and tracking:

Monocular stereo interpretation in own and neighboring lanes (Prometheus 1994); up to 5 vehicles in front and rear hemisphere.

H.1.2 Integrated Transputer System

A.3.6.2c Wheel recognition under oblique views:

Special region-based templates for oblique views from the side (2004)

Video 34 WheelRecognUnderObliqueView 2003

 

 

A.3.6.3 Hybrid Adaptive Cruise Control (1999)

·        Bifocal vision complementing an industrial radar system for more robust distance keeping, VaMP with early version of EMS-vision system;

·        longitudinal control fully autonomous, lateral control by human driver to keep him ‘in the loop’ for better attention.

 

Long-lead-time studies of more demanding vision tasks:

A.3.6.4 Recognition of 3-D shape and occlusion of trucks

·        Approximation of upper body by a rectangular box;

  • determine size parameters from oblique views; early hypothesis generation from partial uncovering of occlusion.

A.3.6.5 Recognition of 3-D vehicle shape and trajectory driven

·        Parameterized vehicle classes from different car types to trucks,

  • stationary camera fixating the car driving on an oval course; Task: Recognize vehicle type and trajectory driven.

 

A.3.6.6 Recognition of humans in traffic scenes

  • Human body modeled with 22 articulated elements;

·        recognize mode of motion (gait) and step frequency & amplitude.

 

A.3.6.7 Detection and recognition of ditches (negative obstacles) while driving off-road on grass surface at speeds up to 10 km/h. EMS-vision including horopter-stereo with ACADIA-board (Sarnoff / PVS), saccadic perception of size of ditch for moving around it.

Complex maneuvering



A.3.7 Turning off onto an unknown crossroad

·        Perception of interception angle, road width and distance,

  • Triggered feed-forward maneuver with superimposed feedback in final phase on the crossroad.

 

 

A.3.8 LongDistanceDrive VaMP 1995 (Munich – Odense, Denmark)

·        Collect statistical data on performance level achieved at the end of the 2nd-generation (black / white, edge-based) dynamic vision system;

  • > 1600 km (~ 95%) driven autonomously (longitudinal & lateral)

 



A.3.9 On-Off-Road Mission performance 2003

Final demonstrations of the capabilities developed with

H.1.3 EMS vision system

On- / off-road driving, transitions, turn-offs, avoid negative obstacles.

M3.0 Functional system integration