Neural Network Paradigm for Classification of Defects on PCB

Abstract

Satu teknik baru dicadangkan untuk mengkelaskan kerosakan yang boleh terjadi pada PCB menggunakan paradigma rangkaian neural. Algoritma untuk membahagi–bahagikan imej menjadi corak primitif, melingkupi corak primitif berkenaan, penandaan corak, normalisasi corak, dan pengkelasan telah dibangunkan berdasarkan pemprosesan imej morfologi penduaan dan rangkaian neural Learning Vector Quantization (LVQ). Ribuan corak rosak telah digunakan untuk tujuan latihan, dan rangkaian neural diuji untuk menilai prestasinya. Satu imej PCB yang rosak digunakan untuk memastikan teknik yang dicadangkan berfungsi. Kata kunci: PCB, pengkelasan kerosakan, pemprosesan imej morfologi, LVQ A new technique is proposed to classify the defects that could occur on the PCB using neural network paradigm. The algorithms to segment the image into basic primitive patterns, enclosing the primitive patterns, patterns assignment, patterns normalization, and classification have been developed based on binary morphological image processing and Learning Vector Quantization (LVQ) neural network. Thousands of defective patterns have been used for training, and the neural network is tested for evaluating its performance. A defective PCB image is used to ensure the function of the proposed technique. Key words: PCB, defects classification, morphological image processing, LVQ

HRIPCB: a challenging dataset for PCB defects detection and classification

“…After getting the locations of defects, the next step is to identify the defect category. Conventional methods are based on pixel‐by‐pixel comparison between template and test image to select enough features to represent defects [1–3, 5], which would have non‐ideal result if the binarisation is in poor condition. Nevertheless, by using an end‐to‐end deep learning model, the image of defect area can be sent to the model as input directly to obtain a classification result, thereby avoiding extracting pixel‐based features from the binary image.…”

“…The effectiveness of this model is evaluated on real data from PCB manufacturing industry and accuracy is compared with previously proposed non‐referential approaches. A new technique that classifies the defects using neural network paradigm is introduced by Rudi Heriansyah et al [5]. Various defective patterns representing corresponding defect types are designed and thousands of defective patterns have been used for training and testing.…”

Detection of PCB Surface Defects With Improved Faster-RCNN and Feature Pyramid Network

“…This method can effectively shorten the detection time to 1s per slice. Rudi Heriansyah et al [12] manually designed various defect patterns representing corresponding defect types for training and testing. The results show the effectiveness of neural network-based defect classification technology.…”

Automatic Misalignment Defects Detection Correction in PCB using SURF (Speed up Robust Features) Technique

“…Heriansyah et al [6] provide a methodology that find out the faults on printed circuit board through referential pixel base method and finally neural network is used for their classification. Such classification is uses binary morphological image processing and Learning Vector Quantization neural network to convert algorithm segments in to basic primitive patterns, pattern assignment, enclosing the primitive patterns as well as patterns normalization.…”