Lane-Based Fusion Track Correction for Vehicle Object Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle and driver assistance systems face challenges in accurately determining the position of fusion tracks, leading to potential false braking or non-braking, which can increase the risk of accidents.
Innovation Solution
An object recognition apparatus and method that utilize a processor to compare the state of a fusion track identified through multiple sensors with the state of a LIDAR track, adjusting the fusion track's position accordingly to improve accuracy and reduce false braking or non-braking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fusion track position is determined through multiple sensors without comparison, then the system operates with simpler processing, but the position accuracy deteriorates leading to false braking or non-braking
Solution Approach 1:
The system obtains both LIDAR track and fusion track, compares their positions, and uses the comparison result to adjust the fusion track position. This feedback mechanism ensures that the fusion track position accurately reflects the actual external object position, resolving the contradiction between position accuracy and system complexity by implementing a targeted comparison and adjustment process.
Solution Approach 2:
The processor acts as an intermediary that receives data from multiple sensors (LIDAR, camera, radar), processes the track information, compares positions, and outputs the adjusted fusion track. This intermediary processing step mediates between the raw sensor data and the final position determination, improving accuracy while maintaining manageable system complexity through structured data flow.
2Reliability
If fusion track position accuracy is improved through comparison and adjustment, then false braking and non-braking frequency reduces, but the processing time and computational load increase
Solution Approach 1:
The system performs preliminary comparison between LIDAR track and fusion track positions before making braking decisions. By pre-comparing the track positions and adjusting the fusion track in advance, the system ensures that subsequent braking decisions are based on accurate position information, thereby improving reliability without adding significant time loss during critical decision-making moments.
3Measurement precision
If the system uses only fusion track from multiple sensors, then the system architecture is simpler, but the position accuracy deteriorates causing increased false braking or non-braking
Solution Approach 1:
The system merges LIDAR track data with fusion track data from multiple sensors (camera, radar, LIDAR) to determine the final fusion track position. By combining these data sources and comparing their positions, the system leverages the strengths of each sensor type to achieve higher position accuracy while maintaining a unified processing architecture that manages complexity through integrated multi-sensor fusion.
Data Source
AI summary
An object recognition apparatus includes a processor. The processor may determine a left line and a right line of a lane on which a vehicle is located, obtain a light detection and ranging (LIDAR) track corresponding to an external object, determine a LIDAR track state corresponding to a position of the LIDAR track relative to the left line or the right line, obtain a fusion track corresponding to the external object and obtained through at least two of a LIDAR, a radar, and a camera, determine a fusion track state corresponding to a position of the fusion track relative to the left line or the right line, move the fusion track in a direction determined based on the LIDAR track state and/or the fusion track state, and output a signal indicating the moved fusion track.