Map-Assisted Object Recognition for Driver Assistance Systems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern driver assistance systems in vehicles often misinterpret environmental data, leading to incorrect object recognition and phantom objects, which can trigger unnecessary reactions and reduce safety and comfort during driving.
Innovation Solution
The method involves using prior knowledge from a map with classified objects to improve object recognition and classification by evaluating recorded environmental data with pattern recognition methods, adjusting the probability of existence and relevance of detected objects, and adapting sensor parameters based on expected objects in the environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If driver assistance systems evaluate surroundings data using pattern recognition methods, then object recognition capability is improved, but false recognition of phantom objects increases
Solution Approach 1:
The system performs preliminary localization of the motor vehicle in the map before evaluating surroundings data. By determining the vehicle's position and orientation in advance, the system can pre-filter and weight sensor data based on expected objects in specific areas, reducing false recognition of phantom objects while maintaining accurate object detection.
Solution Approach 2:
The map serves as an intermediary between the sensor system and the evaluation device. The map provides prior information about the environment that mediates the interpretation of sensor data, helping to distinguish real objects from phantom objects by comparing sensor detections with expected map features.
2Productivity
If the system increases sensitivity to detect more objects, then detection coverage is improved, but phantom objects and false alarms increase
Solution Approach 1:
The system applies different evaluation criteria and weighting factors to different spatial areas based on map information. Areas with high expected object density receive different processing parameters than areas where phantom objects are likely, allowing high detection coverage in relevant areas while suppressing false detections in problematic zones.
Solution Approach 2:
The system dynamically adjusts evaluation parameters such as detection thresholds, confidence levels, and weighting factors based on the vehicle's location in the map and expected environmental conditions. This allows the system to maintain high detection sensitivity where appropriate while reducing false positives in areas prone to phantom objects.
3Reliability
If the system uses map data to filter detected objects, then false alarms are reduced, but detection of unexpected objects may be missed
Solution Approach 1:
The system applies map-based filtering partially rather than completely. It uses map information to weight and prioritize detected objects, reducing confidence in objects that don't match map expectations while maintaining the ability to detect and report unexpected objects that may be equally important for safety.
Solution Approach 2:
The system dynamically adjusts the influence of map data based on contextual factors such as object type, location, and detection confidence. Unexpected objects can trigger map updates, allowing the system to adapt to new environmental conditions while maintaining reliable operation based on established map information.
Data Source
Figure 1
Figure 2a~2b
Figure 3a~3b
AI summary
The invention relates to a method for supporting a driver assistance system (4) in a motor vehicle (50), comprising the following steps: providing a map (7), wherein classified objects (20-1, 20-2, 20-3, 20-4) are stored in the map (7) in associated positions; detecting environment data (8) by means of at least one environment sensor (5) of the driver assistance system (4); evaluating the detected environment data (8) by means of an evaluation device (6) of the driver assistance system (4), wherein the detected environment data (8) is evaluated for object recognition according to the classified objects (20-1, 20-2, 20-3, 20-4) stored in the map (7). The invention also relates to an associated device (1).