Vehicle Environment Map Obstacle Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing environment mapping methods for vehicles, particularly using sensors with low transverse resolution like ultrasonic sensors, lead to loss of important information and the creation of artifacts such as elongated obstacles due to overlapping fields of view, which affects accurate obstacle localization and driver assistance functions like parking and maneuvering.
Innovation Solution
A method that incorporates measurement data from multiple surroundings sensors to determine occupancy probabilities, accounting for the dependencies and spatial relationships between their capture regions, allowing for more accurate and artifact-free filling of the environment map, and updates obstacle probabilities based on combined sensor data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If occupancy probabilities from a single sensor are determined independently and used to update environment map, then the processing is simple and fast, but important information is lost and artifacts such as elongated obstacles arise due to low transverse resolution
Solution Approach 1:
The patent merges occupancy probabilities from multiple sensors by considering their spatial relationships and overlapping fields of view. Instead of processing each sensor independently, the method combines measurements from multiple sensors to determine occupancy probabilities, thereby preserving important information while maintaining processing efficiency through coordinated updates of the environment map.
2Ease of manufacture
If occupancy probabilities are determined independently from each sensor, then the calculation is straightforward, but artifacts such as elongated obstacles are created in the environment map
Solution Approach 1:
The patent implements a feedback mechanism where occupancy probabilities from multiple sensors are cross-validated. The method uses measurements from multiple sensors to verify and refine occupancy probabilities, preventing the creation of artifacts like elongated obstacles while maintaining straightforward calculation through systematic probability updates in the environment map.
3Device complexity
If measurements from multiple sensors are processed independently, then the system complexity is low, but accurate obstacle localization is compromised particularly for parking and maneuvering functions
Solution Approach 1:
The patent segments the processing of multi-sensor data by considering the spatial relationships and overlapping fields of view of individual sensors. Instead of processing all sensors uniformly, the method divides the environment map into regions covered by different sensors and processes occupancy probabilities segment by segment, thereby achieving accurate obstacle localization without excessive system complexity.
Data Source
AI summary
An environment map includes cells, each of which is assigned to portions of the environment of a vehicle and each of which is assigned an obstacle probability that represents the probability that the corresponding portion of the environment is occupied by an obstacle. The vehicle has at least two environment sensors, each of which is designed to provide measurement data on the occupancy of a region of the environment by an obstacle, referred to as an obstacle region, in the respective detection region of the sensor. The measurement data describes obstacle regions which extend over multiple portions of the environment, and the detection regions of the environment sensors at most partly overlap. A method for providing the environment map for the vehicle has the following steps: receiving the measurement data from the at least two environment sensors, the measurement data of a first environment sensor identifying an obstacle region; determining occupancy probabilities for the portions of the environment covered by the identified obstacle region of the measurement data of the first environment sensor on the basis of the measurement data of at least one other environment sensor, wherein an occupancy probability for a portion indicates the probability that the corresponding portion of the environment is occupied by an obstacle; and updating the obstacle probability of the environmental map for at least the portions for which the occupancy probability has been determined.


