Vehicle Sensor Fusion for Merging Split Radar Object Tracks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Radar object detection systems in vehicles often erroneously report multiple objects for a single long vehicle, leading to incorrect activation of active safety features like automatic braking, due to the separation of radar returns from the front and rear of large vehicles, which can result in unintended braking events.
Innovation Solution
A method and system that use a combination of radar and camera data to assess conditions, such as constant longitudinal distance and velocity matching, to determine if a radar track is part of a larger vehicle, thereby discounting the erroneous radar track and considering the vehicle as a single unit for active safety feature control, using a processor to integrate data from both sensors and apply deep learning algorithms for object detection and tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radar data is used to detect objects, then object detection capability is improved, but false detection of single vehicle as multiple objects occurs
Solution Approach 1:
The patent combines radar detection data with camera detection data to merge information about detected objects. When radar detects multiple objects but the camera detects only one, the system merges these detections to determine they represent a single vehicle, thereby reducing false detections while maintaining the radar's high detection capability
Solution Approach 2:
The camera acts as an intermediary verification device between the radar and the active safety feature control system. The camera's detection results are used to validate or refute radar detections, serving as a mediator that resolves the contradiction between radar's high detection precision and its tendency toward false positives
2Reliability
If active safety features are activated based on radar detection, then vehicle safety is improved, but unintended braking events occur due to erroneous detections
Solution Approach 1:
The system implements feedback by using camera detection results to verify radar detections before activating active safety features. This feedback loop prevents erroneous radar detections from triggering unintended braking, while still allowing genuine threats to activate safety features when confirmed by both sensors
Solution Approach 2:
The system performs preliminary verification using the camera before activating active safety features based on radar detection. This preliminary action filters out erroneous detections in advance, ensuring that only verified threats trigger safety features and preventing harmful unintended braking events
Data Source
AI summary
Systems and methods of controlling an active safety feature of a vehicle are provided. The systems and methods receive radar data from a radar device of the vehicle and image data from a camera of the vehicle. Object detection and tracking processes are performed on the radar data and the image data to identify and track objects in an environment of the vehicle. Conditions are assessed with respect to identified objects to ascertain whether a radar track is erroneously reported as a separate object to a camera track. When the conditions are assessed to be true, an object corresponding to the camera track is used as an input for controlling an active safety feature of the vehicle and an object corresponding to the radar track is discounted for controlling the active safety feature of the vehicle.