Multi-Modal Imaging for PCB Defect Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Automated Optical Inspection (AOI) systems for Printed Circuit Boards (PCBs) are limited in their ability to efficiently scan surfaces using multiple illumination configurations during a single sweep, leading to incomplete or inaccurate defect detection due to registration requirements and inefficient processing times.
Innovation Solution
The system employs a method and apparatus that capture a sequence of images using different illumination configurations during a single sweep, with precise shifts between each image to ensure overlap and accurate alignment, allowing for the construction of large area images without the need for extensive registration, using a combination of line and area cameras and high-frequency illumination switching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple illumination configurations are used to detect a wider range of defects, then defect detection capability is improved, but processing time and system complexity increase
Solution Approach 1:
The patent segments the image acquisition process by capturing images with different illumination configurations during a single continuous sweep, rather than performing separate scans for each illumination type. This segmentation of the acquisition process from the processing process allows parallel capture of multiple illumination states without increasing total inspection time
Solution Approach 2:
The system performs preliminary action by capturing all necessary illumination configurations during the initial single sweep, storing these images for later processing. This preliminary capture eliminates the need for repeated sweeps, as all illumination variants are acquired upfront and can be processed sequentially or in parallel without additional time penalty
2Reliability
If multiple illumination configurations are used to detect a wider range of defects, then defect detection capability is improved, but device complexity increases
Solution Approach 1:
The patent implements universality by using a single imaging system that can capture multiple illumination configurations during one sweep. The system is designed to be multi-functional, handling different illumination types (dark field, bright field, polarized, fluorescent) through a unified acquisition and processing framework, eliminating the need for separate dedicated systems for each illumination type
Solution Approach 2:
The system employs parameter changes by varying illumination characteristics (intensity, angle, polarization state, wavelength) during the sweep rather than changing physical hardware configurations. This allows the same imaging system to adapt to different inspection requirements by modifying illumination parameters dynamically, reducing mechanical complexity
3Measurement precision
If images are captured with precise shifts and overlaps during a single sweep, then registration accuracy is improved, but data processing complexity increases
Solution Approach 1:
The system performs preliminary action by establishing precise geometric relationships and overlap patterns during the single sweep acquisition phase. Registration markers and reference features are captured in advance with known spatial relationships, enabling subsequent processing to rely on pre-established geometric constraints rather than complex real-time alignment algorithms
Data Source
AI summary
A method for scanning a surface with a number of different illumination configurations, the method comprises capturing a plurality of images in a sequential manner during a single sweep, each image including one or more lines of pixels, sequentially altering an illumination configuration used to capture the plurality of images according to a predefined sequence of illumination configurations and shifts of the relative position of the imaging unit for capturing each of the plurality of images, and repeating the sequence of illumination configurations settings and associated image capture positions until a desired area of the surface is scanned, wherein said predefined shift is between 10 pixels and less then one pixel.