Vehicle Interior Defect Detection Using Template Image Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Visual inspections by human operators often fail to detect certain vehicle component defects due to vision limitations and environmental factors, leading to incomplete defect detection during manufacturing.
Innovation Solution
A system and method utilizing imaging devices and a controller to generate and compare modified vehicle images with template images, employing cross-correlation, normalized cross-correlation, or sum-of-squared difference values to identify defects, and iteratively adjust image scales and perform segmentation-based processing to determine defect presence and type.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual inspections are performed by human operators, then the inspection process is simple and low-cost, but the detection accuracy and completeness are insufficient due to vision limitations and environmental factors
Solution Approach 1:
The patent replaces the mechanical visual inspection system (human operators) with an automated image processing system that captures images of vehicle interiors, converts them to different color spaces, and compares them against template images to detect defects automatically, thereby eliminating human vision limitations
Solution Approach 2:
The patent transforms the inspection approach by changing the parameter representation of images through color space conversion (RGB to YCbCr, HSV, or LAB), enabling the detection system to identify defects that are not apparent in standard color representations
2Reliability
If automated image processing is implemented, then defect detection accuracy improves, but the processing complexity and computational requirements increase
Solution Approach 1:
The patent segments the complex image processing task into distinct stages: image capture, color space conversion, template matching, and defect identification. This segmentation allows each stage to be optimized independently while maintaining overall system reliability
Solution Approach 2:
The patent introduces color space conversion as an intermediary step between image capture and defect detection, transforming images into alternative color representations that facilitate more reliable defect identification through template comparison
3Measurement precision
If multiple color conversion routines are applied, then defect detection completeness improves, but the processing time and computational load increase
Solution Approach 1:
The patent applies partial action by selectively using one or more color conversion routines (YCbCr, HSV, or LAB) based on the specific inspection requirements, rather than always applying all possible conversions, thus balancing detection completeness with processing efficiency
Data Source
AI summary
A method includes obtaining an image from one or more imaging devices disposed in the interior of the vehicle, modifying the image based on a color conversion routine to generate a source image, generating a conformity value based on a comparison between the source image and a template image corresponding to the interior of the vehicle, where the conformity value is a cross-correlation value, a normalized cross-correlation value, a sum-of-squared difference value, or a combination thereof, and determining a presence of a defect based on the conformity value.