Javascript must be enabled to continue!
Detection method of PCB component based on automatic optical stitching algorithm
View through CrossRef
Purpose
– This paper aims to propose image stitching by reduction of full line and taking line image as registration image to solve the problem of automatic optic inspection in PCB detection. In addition, surf registration was introduced for image stitching to improve the accuracy and speed of stitching.
Design/methodology/approach
– First, image stitching proceeded by method of full line reduction and taking line image as registration image; second, surf registration was introduced based on the traditional PCB image stitching algorithm. Scale space of the image pyramid was adopted for confirming relative future points between stitching image. The registration means of nearest neighbourhood and next neatest neighborhood was selected for feature matching and fused in region of interest to fulfil image stitching.
Findings
– The improved stitching algorithm with small data size of image, high speed and noncumulative transitive error eliminated displacement deviation and solved the stitching gap caused by uneven illumination, to greatly improve the accuracy and speed of stitching.
Research limitations/implications
– The research of this paper can only used for appearance detection and cannot be used for solder joint inspection with circuit detection or invisible solder joint detection; it can identify and mark PCB component defects but cannot classify automatically, thus artificial confirmation and processing is needed.
Originality/value
– Based on the traditional image stitching means, this paper proposed full line reduction for image stitching, which reduces processing of data and speeds up image stitching; in addition, surf registration was introduced into the study of PCB stitching algorithm, which greatly improves the accuracy and speed of stitching and solves stitching gap formed by opposite variation trend of image local edge caused by uneven illumination.
Title: Detection method of PCB component based on automatic optical stitching algorithm
Description:
Purpose
– This paper aims to propose image stitching by reduction of full line and taking line image as registration image to solve the problem of automatic optic inspection in PCB detection.
In addition, surf registration was introduced for image stitching to improve the accuracy and speed of stitching.
Design/methodology/approach
– First, image stitching proceeded by method of full line reduction and taking line image as registration image; second, surf registration was introduced based on the traditional PCB image stitching algorithm.
Scale space of the image pyramid was adopted for confirming relative future points between stitching image.
The registration means of nearest neighbourhood and next neatest neighborhood was selected for feature matching and fused in region of interest to fulfil image stitching.
Findings
– The improved stitching algorithm with small data size of image, high speed and noncumulative transitive error eliminated displacement deviation and solved the stitching gap caused by uneven illumination, to greatly improve the accuracy and speed of stitching.
Research limitations/implications
– The research of this paper can only used for appearance detection and cannot be used for solder joint inspection with circuit detection or invisible solder joint detection; it can identify and mark PCB component defects but cannot classify automatically, thus artificial confirmation and processing is needed.
Originality/value
– Based on the traditional image stitching means, this paper proposed full line reduction for image stitching, which reduces processing of data and speeds up image stitching; in addition, surf registration was introduced into the study of PCB stitching algorithm, which greatly improves the accuracy and speed of stitching and solves stitching gap formed by opposite variation trend of image local edge caused by uneven illumination.
Related Results
Development of electro‐optical PCBs with polymer waveguides for high‐speed intra‐system interconnects
Development of electro‐optical PCBs with polymer waveguides for high‐speed intra‐system interconnects
PurposeThe purpose of this paper is to study fabrication of optical‐PCBs on panel scale boards in a conventional modern PCB process environment. It evaluates impacts on board desig...
Unsupervised Deep Learning for Enhanced holoentropy Image Stitching
Unsupervised Deep Learning for Enhanced holoentropy Image Stitching
Traditional feature-based image stitching technologies rely heavily on feature detection quality, often failing to stitch images with few features or low resolution. The learning b...
A real-time stitching method of aerial images based on tile map
A real-time stitching method of aerial images based on tile map
Aiming at the problems of small speed and large memory resource
consumption for aerial images got by Unmanned Aerial Vehicle (UAV),
which caused by the high pixel, high precision a...
Co-Pyrolysis of PCB with Cotton Stalk: Product Characterisation
Co-Pyrolysis of PCB with Cotton Stalk: Product Characterisation
Abstract
With advancing technology, PCB (Printed circuit board), one of the most important components of e-waste, has become a source of pollution due to an ineffective was...
Abnormal Status Detection of Catenary Based on TSNE Dimensionality
Reduction Method and IGWO-LSSVM Model
Abnormal Status Detection of Catenary Based on TSNE Dimensionality
Reduction Method and IGWO-LSSVM Model
Background:
Catenary is a crucial component of an electrified railroad's traction power supply system.
There is a considerable incidence of abnormal status and failures due to prol...
Contributions to optimal detection in OTDM and OCDMA optical receivers
Contributions to optimal detection in OTDM and OCDMA optical receivers
Recent developments in optical communication systems have increased the performance of optical networks. Low attenuation fiber optics, high spectral purity lasers and optical ampli...
Neural Network Paradigm for Classification of Defects on PCB
Neural Network Paradigm for Classification of Defects on PCB
Satu teknik baru dicadangkan untuk mengkelaskan kerosakan yang boleh terjadi pada PCB menggunakan paradigma rangkaian neural. Algoritma untuk membahagi–bahagikan imej menjadi corak...
Assessment of residual exposure to PCBs in metallurgy
Assessment of residual exposure to PCBs in metallurgy
Objective: To evaluate the occupational exposure to polychlorinated biphenyls (PCBs) in 56 workers employed in 6 electric arc furnace steelmaking plants and 2 secondary aluminum sm...


