Image-Calibrated Traffic Signal Warning for Legacy Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to provide effective safety features for legacy vehicles to detect approaching traffic signals and stop signs without relying on native sensors, due to calibration difficulties and reliance on external data transmission.
Innovation Solution
A system and method using images from a cell phone, dash cam, or other portable electronic device to warn drivers of potential red light violations or failure to stop at stop signs, calibrated using relative sizes of targets in images, and estimating vehicle speed and acceleration without GPS.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If image-based distance calibration is used without native sensors, then safety features can be added to legacy vehicles at low cost, but measurement precision deteriorates due to calibration difficulties
Solution Approach 1:
The patent uses a portable electronic device with camera as an intermediary tool to capture images of traffic signals and stop signs. This intermediary enables distance measurement without requiring native sensors in the vehicle, thus keeping costs low while still achieving functional distance estimation through image analysis and calibration techniques
Solution Approach 2:
The system performs calibration by capturing images at known distances and adjusting measurement parameters accordingly. By changing parameters such as image resolution, field of view, and calibration factors based on captured images, the system optimizes measurement precision for each specific viewing condition, thereby improving accuracy without adding expensive hardware
2Loss of information
If GPS is used to estimate vehicle speed and acceleration, then speed and acceleration can be determined, but reliability deteriorates due to GPS inaccuracies
Solution Approach 1:
The patent replaces GPS-based speed and acceleration determination with image-based optical flow analysis. By analyzing the motion of traffic signals and stop signs across sequential images, the system calculates vehicle speed and acceleration directly from visual data, eliminating reliance on GPS and its associated inaccuracies
Solution Approach 2:
The system creates a visual copy of the environment through sequential images and analyzes the motion of objects within these copies to determine vehicle dynamics. This copying approach allows speed and acceleration to be derived from image sequences rather than direct sensor measurements, improving reliability by using multiple data points over time
3Device complexity
If zoom level changes are not compensated in image analysis, then device complexity is reduced, but measurement precision deteriorates due to altered calculated distances
Solution Approach 1:
The system incorporates feedback by detecting the zoom level from captured images and using this information to adjust distance calculations. The zoom level detection provides feedback about the current optical state, which is then used to compensate for magnification effects in the image analysis algorithm, ensuring accurate distance measurements across different zoom levels
Solution Approach 2:
The patent makes the image analysis system dynamic by automatically adapting to changing zoom levels. Rather than using a fixed analysis approach, the system adjusts its parameters based on the detected zoom state, allowing it to maintain measurement precision across varying optical conditions without requiring complex manual calibration for each zoom level
Data Source
AI summary
An analytic routine uses images from a cell phone, dash cam, or other portable electronic device to warn a driver that his/her vehicle is likely go through a red light, or fail to stop or sufficiently slow down at a stop sign or other stop signal. The analytic routine can be calibrated using relative sizes of a target in the first and second images, and apparent size of the target in the second image when at a time when the device is at an estimated or calculated distance to the target. Following calibration, the calibrated analytic routine uses information obtained from a third image to derive an estimated travel time to a second stop signal, and warns the driver to adjust a speed of the vehicle.