基于卷積神經網絡與機器視覺的紙張塵埃度測量系統(tǒng)的設計與應用研究

打開文本圖片集

作者簡介：李歡，工程師；研究方向：人工智能與機器視覺在檢測設備研發(fā)中的應用、設備精度與自動化水平的提升。

關鍵詞：紙張塵埃度；卷積神經網絡（CNN）；機器視覺；圖像處理中圖分類號：TS77；TP391.41 文獻標識碼：A DOI： 10.11980/j.issn.0254-508X.2025.08.020

Designand Research of Paper Dirt Determination System Based on Convolutional Neural Network andMachineVision

LI Huan*LI Liang （Wuhan Product Quality Testing Institute Co.，Ltd.，Wuhan，Hubei Province，430048） （*E-mail：lihuanWHQT@163.com）

Abstract：Thisstudydesigedapaperdirtdeterinationsystembasedonconvolutionalneuralnetwork （CN）andmachinevisionThesystemwasconstructedwithtwomodules，modeltrainingandtesting.Highresolutionsannerswereusedtoobtaindirtdatasetsandiagesf papersamples.Dieretoptimzatioalgors wereapldtotainteclasificatiomodel，andadagonalmeasurntalgorwas adopted.Asandarddirtpieltableasceaedfgradgandlasifatiosatists，erbycalulatingthdirthsultssodat the precision of the system could reach 0.007mm2 ，which was better than the requirement specified in GB/T 1541—2O13“Paper and board—Determination of dirt".The classification precision reached 95. 89 % ，enabling full-range determination of various paper products.The repeatabilitydetermiationerorofasinglesapleasO.Compadwitmaualdetetiontesingesapledetectiontestigtieofteste was reduced by about 97% ，realizing efficient and accurate detection of dirt in paper products.

ey words： paper dirt;convolutional neural network （CNN）；machine vision;image processin{

紙張塵埃度是判定紙張外觀質量的關鍵指標之一，及時將紙張塵埃度的檢驗結果反饋給生產企業(yè)人員，有助于追溯生產環(huán)節(jié)，改善工藝流程。（剩余8826字）