摘要
在中国制造业加速向智能化转型的宏观背景下,特别是在国家积极倡导并大力推动新质生产力发展的浪潮中,各类电子设备迎来了前所未有的快速发展机遇。随之而来的是,电子产品内部的核心组件——印制电路板(Printed Circuit Board,PCB)的可靠性要求被提升到了新的高度。PCB表面的任何缺陷都将直接影响到电子设备的整体可用性和长期稳定性。然而,传统的依赖于人工手动筛查的方法不仅效率低下,而且极易出现漏检情况,这一瓶颈问题严重制约了企业的生产效率和产品质量。对此,进行了一种基于改进YOLOv10的PCB缺陷检测方法的相关研究,包括引入CBAM注意力机制、优化损失函数以及利用分组卷积对传统目标检测头进行优化等,以提高模型对PCB图像中细微特征的学习能力和检测精度。
Under the macro background of accelerating the intelligent transformation of China's manufacturing industry,especially in the wave where the country is actively advocating and vigorously promoting the development of new quality productivity,all kinds of electronic equipment have ushered in unprecedented rapid development opportunities.As a result,the reliability requirements of Printed Circuit Board(PCB),the core component of electronic products,have been raised to a new level.Any defects on the PCB surface will directly affect the overall availability and long-term stability of the electronic device.However,the traditional method of relying on manual screening is not only inefficient,but also prone to missed detection.This bottleneck problem seriously restricts the production efficiency and product quality of enterprises.In this regard,a PCB defect detection method based on improved YOLOv10 is studied,including introducing CBAM Attention Mechanism,optimizing loss function and optimizing traditional target detection head by group convolution,so as to improve the learning ability and detection accuracy of the model for fine features in PCB images.
作者
何嘉泳
陈芳
张绮婷
陈伟迅
HE Jiayong;CHEN Fang;ZHANG Qiting;CHEN Weixun(Information Engineering Institute,Guangzhou Railway Polytechnic,Guangzhou 511300,China)
出处
《现代信息科技》
2025年第14期27-31,共5页
Modern Information Technology
基金
广州市教育局高校科研项目(2024312419)。
