Enhanced Low-Illumination Image Defect Detection Using Machine Vision
Journal Title: Journal of Industrial Intelligence - Year 2024, Vol 2, Issue 4
Abstract
The detection of image defects under low-illumination conditions presents significant challenges due to unstable and uneven lighting, which introduces substantial noise and shadow artifacts. These artifacts can obscure actual defect points while simultaneously increasing the likelihood of false positives, thereby complicating accurate defect identification. To address these limitations, a novel defect detection method based on machine vision was proposed in this study. Low-illumination images were captured and decomposed using a noise assessment-based framework to enhance defect visibility. A spatial transformation technique was then employed to distinguish between target regions and background components based on localized variations. To maximize the contrast between these components, the Hue-Saturation-Intensity (HSI) color space was leveraged, enabling precise segmentation of low-illumination images. Subsequently, an energy local binary pattern (LBP) operator was applied to the segmented images for defect detection, ensuring improved robustness against noise and illumination inconsistencies. Experimental results demonstrate that the proposed method significantly enhances detection accuracy, as confirmed by both subjective visual assessments and objective performance evaluations. The findings indicate that the proposed approach effectively mitigates the adverse effects of low illumination, thereby improving the accuracy and reliability of defect detection in challenging imaging environments.
Authors and Affiliations
Yan Li
Keywords
Related Articles
- EP ID EP768670
- DOI 10.56578/jii020403
- Views 9
- Downloads 0
Numerical Analysis of Viscosity and Surface Tension on Microdroplet Dynamics in Microelectromechanical Systems Applications
Microelectromechanical systems (MEMS) have instigated transformative advancements, notably in controlled microdroplet generation, offering applications across diverse industrial sectors. Precise control of fluid quantiti...
The Challenges of Integrating AI and Robotics in Sustainable WMS to Improve Supply Chain Economic Resilience
The integration of artificial intelligence (AI) and robotics into the warehouse management system (WMS) has substantially advanced supply chain (SC) operations, offering notable improvements in efficiency, accuracy, and...
Fuzzy Control of Active Vehicle Suspensions for Enhanced Safety in Goods Transport
Suspension systems play a critical role in ensuring the safety, comfort, and stability of vehicles during the transportation of both passengers and goods. Among various suspension technologies, active or electronic suspe...
Navigating Complexity: A Multidimensional Neutrosophic Fuzzy Hypersoft Approach to Empowering Decision-Makers
Urban transportation systems, characterized by inherent uncertainty and ambiguity, present a formidable challenge in decision-making due to their complex interplay of factors. This complexity arises from dynamically shif...
Exploring the Impact of Artificial Intelligence Integration on Cybersecurity: A Comprehensive Analysis
The rapid advancement of technology has correspondingly escalated the sophistication of cyber threats. In response, the integration of artificial intelligence (AI) into cybersecurity (CS) frameworks has been recognized a...