| 注册
首页|期刊导航|电子器件|基于深度学习的电路板色环电阻检测方法

基于深度学习的电路板色环电阻检测方法

胡腾波 黄志亮 黄经州

电子器件2025,Vol.48Issue(2):366-371,6.
电子器件2025,Vol.48Issue(2):366-371,6.DOI:10.3969/j.issn.1005-9490.2025.02.020

基于深度学习的电路板色环电阻检测方法

Circuit Board Color Ring Resistance Detection Method Based on Deep Learning

胡腾波 1黄志亮 2黄经州3

作者信息

  • 1. 日照职业技术学院电子信息工程系,山东 日照 276826
  • 2. 浙江师范大学物理与电子信息工程学院,浙江 金华 321004
  • 3. 杭州电子科技大学自动化学院,浙江 杭州 310018
  • 折叠

摘要

Abstract

Color ring resistor is a kind of electronic component that distinguishes resistance value according to color.It has a high false de-tection rate after being affected by environmental factors such as light.Therefore,a circuit board color ring resistance detection method based on deep learning is proposed.The convolution neural network model is constructed by preprocessing all the collected color ring re-sistance images,and the cross entropy loss function is used to segment the color ring resistance images to obtain the color ring resistance characteristics of the circuit board.The number of color rings is judged according to HSI color model,the extracted color ring features are corrected,and the circuit board color ring resistance is detected by using color quantization method.The experimental results show that the minimum calculation amount and parameter amount of the proposed method are 28.96 GMac and 8.25 M,respectively.The color detection performance of color ring resistance is good,which proves that the proposed method can obtain high-precision detection results.

关键词

深度学习/电路板/色环电阻检测/卷积神经网络/交叉熵损失函数

Key words

deep learning/circuit board/color ring resistance detection/convolution neural network/cross entropy loss function

分类

信息技术与安全科学

引用本文复制引用

胡腾波,黄志亮,黄经州..基于深度学习的电路板色环电阻检测方法[J].电子器件,2025,48(2):366-371,6.

基金项目

浙江省自然科学基金项目(LY18F010017) (LY18F010017)

电子器件

1005-9490

访问量5
|
下载量0
段落导航相关论文