Vehicle Camera Fail-Safe Using Sun Position Geometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fail-safe technologies for camera image recognition in autonomous vehicles do not effectively recognize or promptly determine failures due to sudden changes in illumination, such as when entering a tunnel or changing vehicle direction.
Innovation Solution
A method and apparatus that determine camera image recognition fail-safe by considering the position of the sun, using location, time, and orientation information to generate a straight line between the sun and the camera, and determining if this line intersects the camera's lens surface, thereby identifying potential errors in image recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a pixel segmentation based artificial intelligence learning model is used to determine image recognition errors, then the system can identify failures in camera image information, but the system fails to recognize or delays determination when sudden illumination changes occur (e.g., entering a tunnel or changing vehicle direction)
Solution Approach 1:
The system pre-calculates the sun's position (altitude and azimuth) based on location, date, and time information before image recognition occurs. By determining the sun's position in advance and checking whether it falls within the camera's field of view, the system can immediately identify potential illumination interference without waiting for AI model analysis, thereby eliminating determination delays during sudden illumination changes
Solution Approach 2:
The patent introduces an intermediary geometric calculation method that bridges the gap between environmental conditions (sun position) and image recognition outcomes. By calculating whether a straight line from the sun passes through the camera's lens surface, the system creates an intermediate verification step that quickly identifies illumination-related failures before they affect the AI recognition process
2Reliability
If the system uses AI learning models for fail-safe determination, then it can detect image recognition errors, but it cannot promptly identify failures caused by sudden illumination changes such as entering a tunnel or changing vehicle direction
Solution Approach 1:
The patent replaces the purely AI-based mechanical recognition system with a geometric-calculation-based system. Instead of relying on the AI model to detect illumination-related failures, the system uses mathematical calculations of sun position and camera geometry to directly determine whether illumination interference is occurring, achieving both high accuracy and rapid response
Solution Approach 2:
The system changes the detection parameters from analyzing image content through AI models to calculating environmental parameters (sun altitude and azimuth angles) and geometric relationships. This parameter transformation allows the system to quickly identify illumination-related failures by comparing calculated sun positions with camera field of view boundaries
Data Source
AI summary
Disclosed are embodiments for a method and apparatus for detecting soiling of camera image recognition. In an embodiment, the method includes obtaining an altitude and an azimuth of the sun based on a current date, a current time, and location information of a vehicle, generating a straight line between the sun and a camera in a three-dimensional space based on the altitude and the azimuth of the sun, generating a lens surface of the camera in the three-dimensional space based on a moving direction of the vehicle and a lens surface angle of the camera, and determining that an error occurs in image recognition of the camera when the straight line passes through the lens surface.