Image-Based Facility Position Matching for Road Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for detecting anomalies in road facilities using in-vehicle cameras face challenges with GPS positioning errors, environmental changes, and subject shielding, making it difficult to accurately select reference images for comparison.
Innovation Solution
An information processing apparatus estimates the relative position of objects in images using Structure from Motion (SfM) or stereo matching, generating specifying information that associates object identification with facility information to enable accurate selection of images captured at the same position, independent of GPS accuracy or environmental changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS positioning is used to select reference images, then image matching can be performed, but GPS positioning errors cause inaccurate selection of reference images
Solution Approach 1:
The patent introduces a camera pose estimation system as an intermediary between GPS positioning and reference image selection. Instead of directly using GPS coordinates to match images, the system estimates the camera's position and orientation (pose) relative to road facilities using computer vision techniques. This intermediary step compensates for GPS errors by providing more accurate relative positioning information based on visual features of the road environment.
Solution Approach 2:
The patent replaces the mechanical GPS positioning system with a visual-based camera pose estimation system. Rather than relying on satellite-based mechanical positioning, the system uses image processing and feature matching to determine camera position and orientation. This substitution eliminates GPS positioning errors by using visual cues from the road scene itself for positioning.
2Ease of operation
If images with the same imaging position are selected for comparison, then time-series changes can be easily detected, but environmental changes and subject shielding make it difficult to identify true facility changes
Solution Approach 1:
The patent segments the image analysis process into multiple independent components: road facility detection, camera pose estimation, and change detection. By dividing the task of identifying facility changes from the task of positioning, the system can accurately detect anomalies even when environmental conditions change. The segmentation allows each component to focus on its specific function without interference from other variables.
Solution Approach 2:
The patent changes the parameters used for image comparison from simple positional matching to multi-parameter analysis including camera pose (position and orientation), detected facility positions, and relative geometries. By using these transformed parameters, the system can distinguish between apparent changes caused by different viewing angles and actual facility changes, maintaining high detection accuracy despite environmental variations.
3Quantity of substance
If multiple images are captured from moving vehicles, then more data is available for analysis, but it becomes more difficult to select and compare images captured at the same position
Solution Approach 1:
The patent implements a self-service system where the image processing apparatus automatically performs camera pose estimation and reference image selection without manual intervention. The system extracts features from captured images, estimates camera positions and orientations, and automatically identifies the most suitable reference images for comparison. This automation handles the complexity of processing multiple images, making the system scalable regardless of the number of captured images.
Data Source
AI summary
An information processing apparatus includes one or more hardware processors configured to function as an acquisition unit, a detection unit, an estimation unit, and a generation unit. The acquisition unit acquires a first image and a second image. The detection unit detects an object captured in the first image by using at least the first image. The estimation unit estimates a relative position of the object based on an imaging position of a target image by using the target image that is at least one of the first image and the second image. The generation unit generates specifying information in which object identification information for identifying the object, facility information based on the relative position, and image identification information for identifying the target image are associated with each other.