Radar Vision Fusion Matching Algorithm for Automated Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing object tracking systems face challenges in accurately matching radar and vision tracks due to the complementary strengths and weaknesses of radar and camera sensors, leading to increased complexity and cost in data fusion processes.
Innovation Solution
An object tracking system that employs a radar/vision matching algorithm to assign vision-identifications to camera tracks and radar-identifications to radar-globs, determining match-feasibility based on azimuth, distance, and speed differences, and calculating match-probabilities through low-pass filtering to reorganize associations between vision-tracks and radar-globs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If data fusion is performed to combine radar and camera detection data, then object tracking accuracy is improved, but computing hardware cost and complexity increase
Solution Approach 1:
The patent segments the data fusion process into distinct modules: radar detection module, camera detection module, data association module, and tracking module. Each module handles specific tasks independently, allowing parallel processing and reducing overall system complexity while maintaining tracking accuracy through coordinated operation of specialized components
Solution Approach 2:
The patent introduces an intermediary data association module that acts as a mediator between radar and camera detection modules. This intermediary component performs track initialization, update, and deletion operations, transforming raw detection data into associated tracking information without requiring complex direct integration between sensor modules
2Measurement precision
If comprehensive data fusion is performed to match all radar tracks with all camera tracks, then matching accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by pre-establishing detection thresholds, pre-organizing detection data into structured formats, and pre-defining association rules before the actual matching process. Track initialization operations prepare potential associations in advance, reducing the computational burden during real-time processing
Solution Approach 2:
The patent implements dynamic track management where tracking parameters, association thresholds, and data structures are continuously adapted based on current scene complexity, detection quality, and processing load. This dynamic adjustment allows the system to maintain high matching accuracy while optimizing processing speed according to real-time conditions
3Reliability
If multiple radar tracks are maintained per object, then radar detection coverage is improved, but data processing complexity increases
Solution Approach 1:
The patent merges multiple radar tracks that correspond to the same physical object into a unified track representation during the data association phase. By combining redundant detections and consolidating track information, the system maintains comprehensive radar detection coverage while reducing the number of individual tracks that require separate processing downstream
Data Source
AI summary
An object tracking system suitable for use on an automated vehicle includes a camera, a radar-sensor and a controller. The controller is configured to assign a vision-identification to each vision-track associated with an instance of an object detected using the camera, and assign a radar-identification to each radar-glob associated with an instance of grouped-tracklets indicated detected using the radar-sensor. The controller is further configured to determine probabilities that a vision-track and a radar-glob indicate the same object. If the combination has a reasonable chance of matching it is includes in a further screening of the data to determine a combination of pairings of each vision-track to a radar-track that has the greatest probability of being the correct combination.