ADAS Situational Awareness via Shadow and Lighting Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced driver assistance systems (ADAS) face challenges in achieving situational awareness comparable to human drivers, particularly in anticipating unseen events and objects, due to limitations in classical sensing and perception methods.
Innovation Solution
The method involves acquiring and analyzing image data streams using an image sensor and vision processor to detect objects, shadows, and lighting, and utilizing a situation recognition engine to anticipate traffic situations, including those involving out-of-sight participants, by considering dynamic shadows and artificial lighting, thereby enhancing the system's ability to anticipate and adapt driving maneuvers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If classical sensing and perception methods based on detecting and classifying objects in camera field of view are used, then the system can identify visible objects, but it fails to anticipate unseen events and objects comparable to human driver performance
Solution Approach 1:
The system performs preliminary detection of shadows and lighting conditions before objects fully enter the camera field of view. By analyzing shadow patterns and lighting changes in advance, the system anticipates the approach of unseen traffic participants, enabling earlier reaction time and improved situational awareness without waiting for direct object detection
Solution Approach 2:
The system uses shadows and lighting conditions as intermediary indicators to infer the presence and movement of unseen objects. Instead of directly detecting hidden traffic participants, the system monitors their shadows cast on the ground and lighting changes they cause, which serve as measurable proxies for anticipating unseen events
2Extent of automation
If the system eliminates human supervision for highly automated operations, then automation level increases, but the need for equivalent situational awareness becomes more critical and difficult to achieve
Solution Approach 1:
The system performs preliminary analysis of shadow and lighting data before traffic participants enter the direct field of view, enabling the automated system to anticipate unseen events without human supervision. This advance detection capability allows the highly automated system to maintain situational awareness equivalent to human drivers by preparing responses before objects become visible
Solution Approach 2:
The system continuously monitors shadow patterns and lighting conditions, providing feedback about the approach of unseen traffic participants. This feedback loop enables the automated system to adjust its situational awareness dynamically, maintaining reliable operation at high automation levels by constantly updating its understanding of the environment based on shadow and lighting changes
3Difficulty of detecting and measuring
If the system only detects objects within camera field of view, then detection simplicity is maintained, but anticipation of upcoming events based on indications falls short of human driver performance
Solution Approach 1:
The system uses shadows as intermediary indicators to anticipate unseen traffic participants. By detecting and analyzing shadow patterns cast by objects before they enter the camera field of view, the system maintains relatively simple detection operations while significantly improving event anticipation capability, matching human driver performance in predicting upcoming events
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed is a method for improving situational awareness of an advanced driver assistance system in a host vehicle (1), the method comprising the steps: Acquiring, with an image sensor, an image data stream comprising a plurality of image frames. Analysing, with a vision processor, the image data stream to detect objects, shadows and/or lighting in the image frames. Recognising, with a situation recognition engine, at least one most probable traffic situation out of a set of predetermined traffic situations taking into account the detected objects, shadows and/or lighting. Controlling, with a processor, the host vehicle taking into account the at least one most probable traffic situation. Further disclosed are a corresponding advanced driver assistance system, an autonomous driving system, a corresponding computer program, and a data carrier comprising said computer program.