Enhanced Small Defect Detection on PCBs Using Lightweight YOLOv5s with Hierarchical Clustering and Dynamic Feature Pyramid Network

Author:

Zhou Zhuguo1,Lu Yujun1,Lv Liye1

Affiliation:

1. Zhejiang Sci-Tech University

Abstract

Abstract

An enhanced clustering technique integrated with the YOLOv5s model addresses the challenges of detecting small defect targets on Printed Circuit Boards (PCBs), which are often difficult to locate and prone to high false detection rates. Initially, the method improves the original K-means algorithm by employing a self-developed Hierarchical Density-Based K-means (HDBK-means) algorithm to re-cluster and generate anchor boxes better suited to PCB fault characteristics. Secondly, it replaces the Concentrated-Comprehensive Convolution (C3) module with a novel combination of the Reparameterized Normalized Cross-Stage Partial Efficient Layer Aggregation Network (RepNCSPELAN) module and Spatial and Channel Reconstruction Convolution (SCConv), reducing the model's computational cost without compromising accuracy. Furthermore, the network is enhanced with an adaptive feature selection module to boost its performance in recognizing small targets. Lastly, the GDFPN (Generalized Dynamic Feature Pyramid Network) is used to achieve information interaction across different scales. further enhancing the network's detection accuracy. Comparative studies were conducted on a public PCB dataset. The experimental results demonstrate that the proposed algorithm achieves a mAP (mean Average Precision) of 98.6%, an accuracy of 99.2%, a model size of 10.9M, and an FPS (Frames Per Second) of 138.1. Compared to the original model, the proposed algorithm improves the mAP by 3.8% and the Precision (P) by 2.9%, while reducing the model size by 20.4%, thus fulfilling the requirements for easy deployment.

Publisher

Springer Science and Business Media LLC

Reference25 articles.

1. A path planning algorithm for PCB surface quality automatic inspection;Xiao Z;Journal of Intelligent Manufacturing,2021

2. Laser-induced breakdown spectroscopy (LIBS) applications in the chemical analysis of waste electrical and electronic equipment (WEEE);Costa VC;Trends in Analytical Chemistry,2018

3. LiDAR-only 3D object detection based on spatial context;Qiang W;Journal of Visual Communication and Image Representation,2023

4. A lightweight and efficient one-stage detection framework;Jianchen H;Computers and Electrical Engineering,2023

5. A Two-Stage Multi-Objective Evolutionary Algorithm for Community Detection in Complex Networks;Zhu W;Mathematics,2023

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3