Inspection Device With Image Compositing For Large-Area Defects
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing inspection methods struggle to efficiently detect defects in large areas of vehicles due to the small size of defects relative to the image size, leading to increased inspection time when images are divided, as the number of images to inspect increases.
Innovation Solution
An inspection device that divides captured images into smaller segments, generates composite images by superimposing them, and uses machine learning to determine defects, reducing the size of images inspected and the number of images needed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the captured image is divided into multiple smaller images, then the defect detection accuracy is improved, but the inspection time is increased
Solution Approach 1:
The captured image is divided into multiple first division images, and then these division images are superimposed to create a composite image. This segmentation approach allows the defect detection system to examine smaller regions in detail while maintaining efficiency by processing a single composite image rather than multiple separate images, thus improving defect detection accuracy without proportionally increasing inspection time
Solution Approach 2:
Multiple first division images are superimposed and merged to generate a single composite image that is then used for defect determination. This combining approach consolidates the inspection process into one evaluation step, reducing the total inspection time while preserving the enhanced defect visibility achieved through division
2Productivity
If a larger captured image is used to cover a large area region, then the inspection time is reduced, but the defect detection accuracy is decreased
Solution Approach 1:
The large captured image is segmented into multiple smaller first division images, which are then superimposed to create a composite image. This segmentation enables the system to maintain high defect detection accuracy by examining smaller regions in detail, while the overall inspection efficiency is preserved through the efficient composite image processing approach
Solution Approach 2:
By dividing the large captured image into smaller first division images and superimposing them, the system applies different processing qualities to different regions. The superposition process enhances local defect visibility in the composite image while maintaining coverage of the entire large area, thus achieving both high detection accuracy and efficient large-area inspection
Data Source
AI summary
An inspection device includes: a captured image acquisition unit configured to acquire a captured image obtained by capturing an image of an inspection target; an image division unit configured to divide the captured image which is acquired into a plurality of first division images; a first composite image generation unit configured to superimpose the plurality of first division images to generate a composite image; and a determination unit configured to use the generated composite image and a learning model to determine whether or not a defect is present in the inspection target.