Attached Substance Detection via Edge Intensity Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional attached substance detection devices struggle to detect substances like snow or mud on camera lenses, especially when the entire lens is buried, as they rely on differences in captured images over time, which may not be significant in such cases.
Innovation Solution
The device calculates a score for each divided region of a captured image based on edge intensity, determining if the region is an attached substance area by analyzing pixel regions and their edge intensity thresholds, allowing for early detection even without significant image differences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional techniques use difference between captured images in time sequence to detect attached substances, then the detection method is simple, but the detection timing is delayed and detection accuracy is insufficient when the entire lens is buried
Solution Approach 1:
The image is divided into multiple divided regions, and edge intensity is calculated for each pixel region within these divided regions. This segmentation approach enables detailed local analysis of edge intensity patterns, allowing the system to detect attached substances even when they cover the entire lens, thereby improving detection accuracy without requiring complex additional hardware
Solution Approach 2:
The invention transitions from temporal comparison (comparing images at different times) to spatial analysis (analyzing edge intensity distribution across pixel regions in a single image). By examining the spatial distribution of edge intensities in divided regions, the system can detect attached substances immediately without waiting for temporal changes, thus improving both detection accuracy and timing
2Reliability
If the entire lens is buried by attached substances, then detection is needed, but conventional time-sequence difference methods fail to detect significant changes
Solution Approach 1:
The system performs preliminary analysis by calculating edge intensity in pixel regions of a single captured image before any temporal comparison is needed. By pre-processing the image to extract edge intensity features in divided regions, the system can immediately determine the presence of attached substances without waiting for subsequent images, thus improving detection reliability and eliminating detection delays
Solution Approach 2:
The invention shifts from temporal dimension (comparing across time) to spatial dimension (analyzing within a single image). By examining the spatial distribution of edge intensities across divided regions in one image, the system can reliably detect attached substances covering the entire lens without temporal changes, improving both reliability and timing
Data Source
AI summary
An attached substance detection device according to an embodiment includes a calculation unit and a determination unit. The calculation unit calculates a score of a divided region that is provided by dividing a predetermined target region, based on an edge intensity in each pixel region of a captured image that is captured by an image-capturing device. The determination unit determines whether or not the target region is an attached substance region where an attached substance is attached to a lens of the image-capturing device, based on the score that is calculated by the calculation unit.