Traffic Sign Determination Using Map Data and Image Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic sign determination devices are unable to accurately notify vehicle occupants of speed limit values when no traffic sign is detected from the picked-up image, leading to a low probability and short duration of successful notifications, and a high likelihood of errors in such notifications.
Innovation Solution
A traffic sign determination device that uses image recognition to detect traffic signs, combines this data with map data to assess the adequacy of cooperation between previous and current road links, and considers factors like regulatory speeds and road types to determine the final traffic sign to be notified, even when no current sign is detected, thereby reducing errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the device uses only image recognition to detect traffic signs, then the detection is simple and fast, but the notification probability is low when no sign is detected and error rate is high
Solution Approach 1:
The patent combines image recognition results with map data to determine final traffic signs. When no traffic sign is detected by image recognition or the detection reliability is low, the system merges map data (including road link information and traffic sign information) with the image recognition results to determine the final traffic sign to be notified, thereby increasing notification probability and reducing errors.
Solution Approach 2:
The patent introduces map data as an intermediary to bridge the gap when image recognition fails or provides low-confidence results. The map data serves as a mediator that provides alternative traffic sign information based on road link data, ensuring continuous and reliable traffic sign notification even when direct image detection is unavailable.
2Measurement precision
If the device uses both image recognition and map data, then the accuracy improves, but the processing complexity increases
Solution Approach 1:
The patent applies partial action by selectively using map data only when necessary - specifically when no traffic sign is detected by image recognition or when the detection reliability is low. The system does not always process both image data and map data simultaneously, but rather uses map data as a supplement only when needed, thereby improving accuracy while controlling processing complexity.
Solution Approach 2:
The patent changes the parameter of data source selection based on detection reliability. When image recognition confidence is high, only image data is used; when confidence is low or no detection occurs, the system switches to using map data. This dynamic parameter change optimizes the balance between accuracy and processing complexity.
3Duration of action of moving object
If the device notifies traffic signs based on map data when image detection fails, then notification continuity improves, but error probability increases
Solution Approach 1:
The patent implements feedback by continuously monitoring image recognition results and comparing them with map data. When image recognition detects a traffic sign, the system verifies it against map data and uses the consistent result for notification. This feedback mechanism ensures that notifications based on map data are cross-validated with image recognition when available, maintaining high accuracy while ensuring continuous notification.
Data Source
AI summary
An image recognition section detects a first traffic regulation sign from a picked-up image of an area in front of a vehicle. An acquisition section uses a position of the vehicle to acquire a road link, including a second traffic regulation sign at the position of the vehicle, from a storage section. When no current first traffic regulation sign is detected on a current road link after turning-right or turning-left of the vehicle, a determination section determines a current final traffic regulation sign on the current road link, which is to be notified to a vehicle occupant, by use of a degree of adequacy of cooperation between previous first and second traffic regulation signs detected and acquired on a previous road link before the turning-right or turning-left of the vehicle, but also previous and current factors acquired on the previous and current road links.