Vehicle Obstacle Detection Fusion for Precision-Recall Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in ensuring safety by efficiently avoiding both known and unknown obstacles while minimizing processing power consumption and duplicate detection.
Innovation Solution
A system and method utilizing high precision and high recall detection techniques, combining data from various sensors like RADAR, LIDAR, and cameras to identify and track known objects, and filter out unknown objects, enabling efficient trajectory planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high precision detection is used to identify known objects, then detection accuracy is improved, but processing time increases
Solution Approach 1:
The detection system is segmented into two independent pipelines: high precision detection for known objects and high recall detection for unknown objects. This segmentation allows each pipeline to be optimized independently, with high precision detection focusing on accuracy for classified objects and high recall detection focusing on speed for unclassified objects, thereby resolving the contradiction between detection accuracy and processing time.
2Reliability
If comprehensive sensor data is processed to ensure safety, then detection reliability is improved, but energy consumption increases
Solution Approach 1:
The system applies partial processing by dividing sensor data into two categories: data requiring high precision analysis (for known objects) and data requiring high recall analysis (for unknown objects). This partial action approach ensures that full processing power is not continuously applied to all data, reducing energy consumption while maintaining detection reliability through targeted analysis of critical data segments.
3Adaptability or versatility
If high recall detection is used to identify unknown objects, then detection coverage is improved, but false positive rate increases
Solution Approach 1:
The system introduces an intermediary classification mechanism where detected objects are first evaluated by high recall detection, then routed to appropriate processing paths. Objects identified as known types undergo precision verification, while truly unknown objects proceed to high recall processing. This intermediary classification acts as a mediator that reduces false positives by filtering out known objects before they can be misclassified as unknown anomalies.
4Reliability
If duplicate detection is performed across multiple sensors, then detection reliability is improved, but processing complexity increases
Solution Approach 1:
The system extracts and removes duplicate detections by comparing detection results across multiple sensors and time steps. Detected objects are tracked and matched across sensor inputs, with duplicates identified and eliminated before further processing. This extraction of redundant information maintains detection reliability through multi-sensor verification while reducing processing complexity by eliminating redundant computational paths.
Data Source
AI summary
This disclosure provides systems and methods for controlling a vehicle based on a combination of high precision detection and high recall detection. The disclosed systems and methods can efficiently generate trajectories by reducing duplicate detection or duplicate calculation of objects or obstacles of common and known object types and objects or obstacles without class identification.