Road Sign Detection Using Map Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for detecting road signs from in-vehicle camera images suffer from inaccuracies due to camera performance and environmental conditions, leading to erroneous detections.
Innovation Solution
An information processing apparatus that combines image processing using a machine-learned detection model with map information to determine the presence of road signs, estimating vehicle direction and position to enhance detection accuracy when image reliability is low.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If road sign detection is performed using only image processing from in-vehicle camera, then the system is simple and fast, but detection accuracy deteriorates due to camera performance limitations and environmental conditions
Solution Approach 1:
The patent merges image processing results with map information to determine road sign presence. The determination processing unit combines the reliability from image processing with positional data from map information, creating a hybrid detection system that leverages both visual and spatial data sources to improve accuracy while maintaining reasonable system complexity.
Solution Approach 2:
Map information serves as an intermediary between the image processing unit and the final determination. When image reliability is insufficient, the map information acts as a mediator to confirm or supplement the detection result, resolving the contradiction by providing an additional verification layer without requiring complete system redesign.
2Reliability
If map information is used to determine road sign presence when image reliability is low, then detection accuracy improves, but system complexity and processing time increase
Solution Approach 1:
The system applies map information verification only partially - specifically when the image processing reliability falls below a predetermined threshold. This selective application of the secondary verification method avoids unnecessary processing time for high-confidence detections while providing enhanced reliability for uncertain cases.
Solution Approach 2:
The detection approach dynamically adapts based on image reliability. The system switches between relying primarily on image processing (when reliable) and incorporating map information (when reliability is low), creating a flexible detection strategy that optimizes the balance between speed and accuracy based on real-time conditions.
3Productivity
If road sign detection relies solely on image processing, then processing is fast, but erroneous detections occur due to environmental conditions and camera limitations
Solution Approach 1:
The system implements feedback through the reliability threshold mechanism. When image processing produces low reliability results, the system feedbacks to the determination processing unit to consult map information, creating a closed-loop verification process that reduces erroneous detections while maintaining fast processing for high-confidence cases.
Data Source
AI summary
An information processing apparatus includes an image processing unit configured to detect a road sign from an image taken by an in-vehicle camera and configured to calculate a reliability indicating a probability that the road sign is of a specific type, and a determination processing unit configured to determine presence or absence of the road sign of the specific type in the image, in a case of the reliability less than a threshold value, by referring to map information indicating a position of the road sign of the specific type on a map based on a position at which the image is taken.