scholarly journals Comparative Study of Image Processing and Transfer Learning Techniques for an Automated PCB Fault Detection System

2021 ◽  
Vol 9 (VI) ◽  
pp. 5090-5097
Author(s):  
Muskaan Rajput
Keyword(s):  
Image Processing ◽  
Fault Detection ◽  
Comparative Study ◽  
Transfer Learning ◽  
Printed Circuit Board ◽  
Detection System ◽  
Circuit Board ◽  
Inspection System ◽  
Image Subtraction ◽  
Large Numbers

Printed circuit board (PCB) is one of the most crucial components in most electronic devices. PCBs are manufactured in large quantities, and therefore, maintaining the quality of such large numbers of PCBs is important. An automated inspection system can help in the aspect of quality maintenance. Such a system is able to overcome the limitations of manual inspection for a large number of PCBs. It provides fast detection of defects and hence can prove to be an asset in the manufacturing process. This project aims to achieve fault detection of bare PCBs through two different methods; the traditional algorithmic approach using image processing, which makes use of image subtraction method, and a transfer learning approach, which involves the pre-trained CNN VGG16 model. A comparative study of both the methods is done.

Defects Detection of Printed Circuit Board Based on the Machine Vision Method

2014 ◽  
Vol 494-495 ◽  
pp. 785-788 ◽  
Author(s):  
Wen Bin Wang ◽  
Dao Yuan Liu ◽  
Yu Qin Yao
Keyword(s):  
Image Processing ◽  
Image Registration ◽  
Real Time ◽  
Printed Circuit Board ◽  
Detection Efficiency ◽  
Detection System ◽  
Circuit Board ◽  
Algorithm Analysis ◽  
Real Time Processing ◽  
Morphological Image Processing

The processing of target image using image processing technology, can realize the non-contact online detecting circuit board, thus greatly improve the detection efficiency, reducing the defective rate. This paper provides the detection system based on the methods of pre-processing the standard circuit board image and the circuit board image to be detected, two value segmentation, morphological image processing, image registration and poor shadow detection processing, among them ,image registration is the key. In order to improve the processing speed to achieve real-time processing, image registration using rapid processing algorithm. Analysis of the experimental results, the method can detect the defects on the circuit board to be detected accurately, and can achieve the automatic real-time detection purposes.

PCB welding spot detection with image processing method based on automatic threshold image segmentation algorithm and mathematical morphology

Circuit World ◽  
2016 ◽  
Vol 42 (3) ◽  
pp. 97-103 ◽  
Author(s):  
Shi yuan Wang ◽  
Yang Zhao ◽  
Liuying Wen
Keyword(s):  
Image Processing ◽  
Printed Circuit Board ◽  
High Efficiency ◽  
Detection Efficiency ◽  
Processing Method ◽  
Circuit Board ◽  
Content Type ◽  
Image Processing Method ◽  
Spot Detection ◽  
Welding Spot

Purpose To enhance the detection efficiency of printed circuit board (PCB) soldered dot, this study aims to detect PCB soldered dot with image processing method. Design/methodology/approach Through a series of image processing algorithms, threshold segmentation and feature extraction of the solder joint images were completed; then, sphericity was confirmed based on the area and perimeter, as well as the shape parameter and eccentricity ratio of the calculated region, paving the way for defect pattern recognition. Findings It proves that the method with high efficiency and precision can satisfy the requirements of PCB soldered dot detection. Originality/value This paper is original in presenting a method for PCB welding spot detection.

scholarly journals Implementasi Adaptive Neural Fuzzy Inference System (ANFIS) Untuk Menguji Kualitas Printed Circuit Board (PCB)

2021 ◽  
Vol 5 (2) ◽  
pp. 79-90
Author(s):  
Tatang Rohana Cucu
Keyword(s):  
Visual Inspection ◽  
Printed Circuit Board ◽  
Fuzzy Inference ◽  
Circuit Board ◽  
Inspection System ◽  
Automated Visual Inspection ◽  
Inference System ◽  
Printed Circuit ◽  
Bahasa Indonesia ◽  
Neural Fuzzy

Pengujian kualitas menggunakan teknik pengolahan citra dan kecerdasan tiruan banyak diterapkan dalam berbagai industri, misalnya industri tekstil, perakitan kendaraan, makanan, minuman, perakitan elektronik, dan lain – lain. Pengujian model ini sering disebut dengan istilah Automated Visual Inspection System (AVIS) atau dalam bahasa Indonesia Sistem Inspeksi Visual Otomatis (SIVO). Penelitian ini mengacu pada model sistem inspeksi, di mana objek pengujiannya adalah keping Printed Circuit Board (PCB). Banyak penelitian tentang pengujian PCB yang sudah dilakukan, tetapi masih banyak yang belum memberikan hasil yang optimum, diantaranya waktu akses yang masih lambat, keakuratan data masih rendah, dan tingkat kesalahan yang masih tinggi. Berdasarkan hasil penelitian dan pengujian yang sudah dilakukan, model ANFIS sangat layak dijadikan sebagai model inferensi kecerdasan buatan dalam sistem yang berbasis inspeksi otomatis khususnya menguji kualitas keping PCB, karena terbukti model ANFIS dengan model hybrid trapesium mf memiliki tingkat kesalahan yang sangat kecil yaitu 4.0186e-007 dan untuk tingkat akurasi pengujian datanya mencapai 99%. 

An Automatic Inspection System for Chip Electronic Parts on a Printed Circuit Board

CIRP Annals ◽  
1987 ◽  
Vol 36 (1) ◽  
pp. 399-402 ◽  
Author(s):  
Tadanori Komatsu ◽  
Shinichi Uno ◽  
Mitsuji Inoue ◽  
Shingo Sekiguchi ◽  
Akira Kobayashi
Keyword(s):  
Printed Circuit Board ◽  
Circuit Board ◽  
Inspection System ◽  
Automatic Inspection ◽  
Printed Circuit ◽  
Electronic Parts

An Evaluation of Blink Inspection

1980 ◽  
Vol 22 (2) ◽  
pp. 201-210 ◽  
Author(s):  
J. G. Liuzzo ◽  
C. G. Druryl
Keyword(s):  
Heart Rate ◽  
Printed Circuit Board ◽  
Stopping Time ◽  
Search Time ◽  
Circuit Board ◽  
Inspection System ◽  
Circuit Boards ◽  
Student Subjects ◽  
Inspection Task ◽  
Consistent Tendency

It has been suggested that performance in an industrial inspection task can be improved by rapidly alternating images of the item to be inspected and a perfect item onto the visual field of the inspector. Such a blink inspection system was evaluated for a complex item (a printed circuit board which could contain three flaws) and a simple item (a blank field which could contain one flaw) using eight student subjects for each of five inspection conditions. Three groups, one using blink inspection and two control groups, inspected circuit boards; two groups (one blink and one control) inspected blank fields. Search time was improved by blink inspection for the circuit boards and with different defects showing different degrees of improvement. Only one type of defect showed both reduced search times and reduced errors. For blank fields, performance was worse under blink inspection. Subjects showed a consistent tendency to terminate their search for defects in a logical manner, with stopping time related to search difficulty. Measures of heart rate mean and variability were taken as a pilot study to measure task stress. There was little evidence from the heart rate measures that the task was stressful.

Microfocus X-ray printed circuit board inspection system

Optik ◽  
2014 ◽  
Vol 125 (17) ◽  
pp. 4929-4931 ◽  
Author(s):  
Yu Wang ◽  
Mingquan Wang ◽  
Zhijie Zhang
Keyword(s):  
Printed Circuit Board ◽  
Circuit Board ◽  
Inspection System ◽  
X Ray ◽  
Printed Circuit
Sign in / Sign up

Export Citation Format

Share Document