[1]彭 治,刘 杨,杜永萍,等.基于迁移学习的多场景垃圾图像分类方法[J].计算机技术与发展,2022,32(05):106-111.[doi:10. 3969 / j. issn. 1673-629X. 2022. 05. 018]
　PENG Zhi,LIU Yang,DU Yong-ping,et al.Multi-scene Garbage Image Classification Method Based on Transfer Learning[J].,2022,32(05):106-111.[doi:10. 3969 / j. issn. 1673-629X. 2022. 05. 018]
点击复制

基于迁移学习的多场景垃圾图像分类方法()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年05期
页码:
106-111
栏目:
应用前沿与综合
出版日期:
2022-05-10

文章信息/Info

Title:
Multi-scene Garbage Image Classification Method Based on Transfer Learning
文章编号:
1673-629X(2022)05-0106-06
作者:
彭 治1 ; 刘 杨1 ; 杜永萍1 ; 常燕青2 ; 韩红桂1
1. 北京工业大学 信息学部,北京 100124;
2. 江苏省固体废弃物处理环保装备工程技术研究中心,江苏 常州 213000
Author(s):
PENG Zhi1 ; LIU Yang1 ; DU Yong-ping1 ; CHANG Yan-qing2 ; HAN Hong-gui1
1. Faculty of Information Technology,Beijing University of Technology,Beijing 100124,China;
2. WELLE Environmental Group Co. ,Ltd,Changzhou 213000,China
关键词:
图像识别; 迁移学习; 残差网络; RexNeXt; 垃圾分类
Keywords:
image recognition; transfer learning; residual network; ResNeXt; garbage classification
分类号:
TP391
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 05. 018
摘要:
垃圾分类回收是资源循环再生链条中重要的一环,传统的人工分类方法效率低,而基于神经网络模型的图像识别技术是实现智能化分类回收的重要策略,具有准确度高、泛化性能好等优点。 针对生活垃圾场景的多样性以及垃圾图像数据集缺乏的问题,该文在基于单一背景和复杂背景两种不同垃圾图像场景下,提出一种基于预训练 ResNeXt101 模型进行迁移学习的方法。 利用大规模 ImageNet 图像数据集训练模型参数,并将参数迁移到垃圾图像数据分类模型中进行训练,解决缺乏大规模标注数据的垃圾图像数据集的分类问题,实现玻璃和纸类等 4 种可回收物的分类识别以及厨余垃圾中杂质的检测。 实验结果表明,利用该模型在真实背景下的垃圾图像分类准确率可以达到 87. 42% ,单一背景下分类准确率达到 96. 19% ,实现了高精度的可回收物识别分类以及厨余垃圾中高效的杂质检测,同时有效提升了模型训练速度。
Abstract:
Garbage classification and recycling is an important part of the resource recycling chain. Traditional classification depends on the manual intervention mostly,which is inefficient,while image recognition technology based on neural network model is an important strategy to realize intelligent classification and recycling with the advantages of high accuracy and good generalization performance. In response to the diversity? ? ?of domestic waste scenes and the lack of waste image datasets,a transfer learning method based on a pre-trained ResNeXt101 model with Image Net image dataset is proposed to solve the problem of classifying and recognizing four categories of recyclable household garbage,? ? such? as glass and paper,and detecting whether impurity is mixed in the kitchen garbage under two different image scenarios, including a single background and a complex background. The experimental results show that the garbage image classification based on this model can achieve? 87.42% accuracy in the complex background and 96.19% accuracy in the flat back ground,achieving high accuracy in the identification and classification of recyclables,and efficient detection of impurities in kitchengarbage,while further the training speed of the model is improved.

相似文献/References:

[1]王明平 宋丽梅.基于计算机视觉的车架号采集系统[J].计算机技术与发展,2008,(04):239.
　WANG Ming-ping,SONG Li-mei.Vehicle Identify Number Acquisition System Based on Machine-Vision[J].,2008,(05):239.
[2]徐仕玲 赵敏 徐建波.野外早期火灾图像识别方法研究[J].计算机技术与发展,2008,(06):214.
　XU Shi-ling,ZHAO Min,XU Jian-bo.Research for Early Fire Image Recognition Technology in Outdoors[J].,2008,(05):214.
[3]张婷 吴元君 黄俊 吴建国.选票选举系统中选票图像的预处理方法研究[J].计算机技术与发展,2007,(04):225.
　ZHANG Ting,WU Yuan-jun,HUANG Jun,et al.Research of Image Pre- Processing in Vote Processing System[J].,2007,(05):225.
[4]吴忠 朱国龙 黄葛峰 吴建国.基于图像识别技术的手写数字识别方法[J].计算机技术与发展,2011,(12):48.
　WU Zhong,ZHU Guo-long,HUANG Ge-feng,et al.Handwritten Digit Recognition Based on Image Recognition System[J].,2011,(05):48.
[5]汪晨 张涛 林为民 邓松 时坚 李伟伟.图像识别综述及在电力信息安全中的应用研究[J].计算机技术与发展,2012,(04):161.
　WANG Chen,ZHANG Tao,LIN Wei-min,et al.Image Recognition Review and Application Research in Electric Power Information Security[J].,2012,(05):161.
[6]王珂雅 邱力军.一种新的脑部CT图像异常检测算法[J].计算机技术与发展,2012,(05):185.
　WANG Ke-ya,QIU Li-jun.A New Anomaly Detection Algorithm for Brain CT Image[J].,2012,(05):185.
[7]周诚诚 张代远[].利用图像识别技术过滤海量可疑钓鱼网站[J].计算机技术与发展,2012,(11):246.
　ZHOU Cheng-cheng,ZHANG Dai-yuan.Using Image Recognition Technology to Filter Mass Suspicious Phishing Sites[J].,2012,(05):246.
[8]李 勇,刘战东,张海军.跨项目软件缺陷预测方法研究综述[J].计算机技术与发展,2020,30(03):98.[doi:10. 3969 / j. issn. 1673-629X. 2020. 03. 019]
　LI Yong,LIU Zhan-dong,ZHANG Hai-jun.Review on Cross-project Software Defects Prediction Methods[J].,2020,30(05):98.[doi:10. 3969 / j. issn. 1673-629X. 2020. 03. 019]
[9]毕 杨,宋 飞,王 轩.一种航空维修工具智能管理系统的设计与实现[J].计算机技术与发展,2020,30(07):199.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 041]
　BI Yang,SONG Fei,WANG Xuan.Design and Implementation of an Intelligent Management System of Aviation Maintenance Tools[J].,2020,30(05):199.[doi:10. 3969 / j. issn. 1673-629X. 2020. 07. 041]
[10]严一鸣[],郭星[]. 基于计算机视觉的交互式电子沙盘系统研究[J].计算机技术与发展,2017,27(06):195.
　YAN Yi-ming[],GUO Xing[]. Investigation on Interactive Electronic Sand Table System with Computer Vision[J].,2017,27(05):195.
[11]俞圳韬,刘万里,杨晓辉,等.基于贝叶斯优化和迁移学习的 CNN 算法研究[J].计算机技术与发展,2022,32(S2):68.[doi:10. 3969 / j. issn. 1673-629X. 2022. S2. 012]
　YU Zhen-tao,LIU Wan-li,YANG Xiao-hui,et al.Research on CNN Algorithm Based on BayesianOptimization and Transfer Learning[J].,2022,32(05):68.[doi:10. 3969 / j. issn. 1673-629X. 2022. S2. 012]

更新日期/Last Update: 2022-05-10