Image Processing Apparatus Shadow Reflection Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques for defect detection on inspection target objects face challenges due to complex three-dimensional shapes and significant surface properties, leading to reduced detection accuracy caused by shadow regions and surface reflection.
Innovation Solution
An image processing apparatus and method that acquire multiple captured images of an inspection target object under varying illumination conditions, allowing for both a first inspection based on a normal image and a second inspection based on the captured images themselves, tailored to specific regions affected by shadow and reflection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single inspection method based on normal images is used, then the inspection process is simple, but detection accuracy is reduced in shadow and reflection regions
Solution Approach 1:
The inspection process is segmented into two distinct methods: a first inspection method using normal images for general areas, and a second inspection method using captured images for shadow and reflection regions. This segmentation allows each method to be optimized for its specific use case, improving overall detection accuracy while managing complexity through structured division of labor
Solution Approach 2:
Different inspection methods are applied to different regions of the inspection target object based on local characteristics. The second inspection method is specifically applied to regions identified as shadow or reflection areas, while the first method handles other regions. This local adaptation ensures optimal detection accuracy for each region's specific challenges
2Measurement precision
If multiple inspection methods are employed, then detection accuracy in shadow and reflection regions is improved, but processing complexity increases
Solution Approach 1:
Shadow and reflection regions are identified in advance through image analysis before the inspection process begins. This preliminary identification allows the system to pre-determine which regions require the second inspection method, enabling more accurate surface gradient estimation in problematic areas without adding complexity to the overall processing flow
Solution Approach 2:
The identification of shadow and reflection regions acts as an intermediary step that bridges the simple first inspection method and the more sophisticated second inspection method. This intermediary analysis enables the system to selectively apply the appropriate inspection method based on regional characteristics, improving accuracy while managing processing complexity
Data Source
AI summary
An image processing apparatus is provided. The apparatus acquires a plurality of captured images acquired by shooting an inspection target object under a plurality of illumination conditions. The apparatus performs a first inspection for inspecting a surface profile of the inspection target object based on a normal image indicating a normal direction of each position of the inspection target object. The normal image is generated from the plurality of captured images. The apparatus performs a second inspection for inspecting a surface profile of the inspection target object based on the plurality of captured images, the second inspection being different from the first inspection.