Synchronized Visible-Light Defect Inspection Under Pose and Lighting Variation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Industrial imaging systems face challenges in accurately detecting defects due to part and image sensor pose variations, material property variations, image background variations, and ambient or controlled lighting variations, which impact the effectiveness of classification algorithms.
Innovation Solution
A method and system that utilize a controlled lighting device to project a predefined dynamic lighting pattern onto an object, capture image frames, process them to generate intensity images, and apply machine learning-based classification to assign defect labels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional imaging systems are used for defect detection, then the system structure is simple, but the inspection accuracy is reduced due to part pose variations, material property variations, lighting variations, and background variations
Solution Approach 1:
The patent employs dynamic lighting patterns that change over time to illuminate the object from multiple virtual angles. The lighting device projects sequential lighting patterns that dynamically interact with surface defects, enabling the detection system to capture defect information under varying lighting conditions without requiring physical movement of the object or camera, thus improving inspection accuracy while maintaining relatively simple system structure.
Solution Approach 2:
The lighting device projects a series of periodic lighting patterns onto the object surface. By capturing images during these periodic lighting variations, the system can differentiate between static background elements and dynamic defect reflections, thereby improving defect detection accuracy without significantly increasing system complexity.
2Measurement precision
If dynamic lighting patterns are projected to improve defect detection, then inspection accuracy is enhanced, but the complexity of lighting control increases
Solution Approach 1:
The lighting device divides the illumination into multiple discrete lighting patterns, each designed to highlight specific defect characteristics. By segmenting the lighting into distinct patterns that can be sequentially projected, the system achieves comprehensive defect detection while maintaining manageable control complexity through modular pattern design.
Solution Approach 2:
The system changes lighting parameters such as intensity, angle, and pattern configuration to optimize defect detection. By varying these parameters in a controlled manner, the system enhances defect visibility and detection accuracy without requiring overly complex lighting control mechanisms, as the parameter changes follow predefined patterns.
3Measurement precision
If multiple image frames are captured and processed to generate intensity images, then defect detection capability is improved, but the processing time and computational load increase
Solution Approach 1:
The system extracts only the relevant intensity information from multiple captured image frames by comparing pixel intensities across frames. This extraction approach focuses computational resources on calculating intensity differences and variations that indicate defects, rather than processing all image data, thereby improving defect detection capability while reducing overall processing time.
Solution Approach 2:
The system performs preliminary processing on captured image frames by immediately calculating intensity values and comparing them against reference patterns. This preliminary action prepares the data in advance for final defect analysis, reducing the computational burden during the critical detection phase and minimizing overall processing time.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances inspection accuracy by minimizing the impact of variations and effectively detecting defects in manufactured parts, including both surface and internal flaws.
Implementation Method 1
The controlled lighting device is configured to generate a sequence of varying images for illuminating the object
Data Source
AI summary
System and method that includes: projecting a dynamic lighting pattern from a controlled lighting device towards a manufactured object; capturing a light response of the object, over an inspection period, to the dynamic lighting pattern and generating a set of image frames representing the captured lighting response; processing the set of image frames to generate an intensity image that is based on intensity information across a plurality of the image frames; assigning a quality label based on the generated intensity image.