Abstract:
There is a vast development in manufacturing and packaging technologies, electronic products and equipment now regularly use a variety of precision circuits with large si...Show MoreMetadata
Abstract:
There is a vast development in manufacturing and packaging technologies, electronic products and equipment now regularly use a variety of precision circuits with large sizes and fine features. At the moment, large-scale circuits have extremely small pitches and line widths that can be several microns. Inspection of printed circuit board faults is thus one of the essential components to offer quality assurance. Assuring PCB quality includes checking for mouse bites, missing holes, spur on routing path, open circuit paths, and short circuit paths. Manual inspection takes a lot of time, and since human beings are limited, manual errors are possible because of the need for manual interventions. We are attempting to automate and construct an intelligent visual inspection system using deep neural networks and computer vision. Vision-based analysis of the micro dimensions of a PCB is difficult since a camera can only capture a single view of an image with a limited field of view. To get a wide field of view in this instance, we used picture stitching by registration and fusion. Therefore, this research provides a high-precision micro vision image stitching technique using computer vision and fault detection of PCB using deep learning methodology.
Date of Conference: 26-27 August 2022
Date Added to IEEE Xplore: 12 January 2023
ISBN Information: