《计算机应用》唯一官方网站 ›› 2026, Vol. 46 ›› Issue (3): 940-949.DOI: 10.11772/j.issn.1001-9081.2025030368

• 多媒体计算与计算机仿真 • 上一篇    下一篇

基于多层预处理的城市固废焚烧状态识别方法

张健1(), 于剑波1, 汤健2   

  1. 1.南京信息工程大学 计算机学院、网络空间安全学院,南京 210044
    2.北京工业大学 信息科学技术学院,北京 100124
  • 收稿日期:2025-04-07 修回日期:2025-06-16 接受日期:2025-06-25 发布日期:2025-07-15 出版日期:2026-03-10
  • 通讯作者: 张健
  • 作者简介:于剑波(2001—),男,江苏盐城人,硕士研究生,CCF会员,主要研究方向:机器学习、图像分类
    汤健(1974—),男,辽宁北票人,教授,博士生导师,博士,主要研究方向:小样本数据建模、固废处理过程智能控制。
  • 基金资助:
    国家自然科学基金资助项目(62073006)

Municipal solid waste incineration state recognition method based on multilayer preprocessing

Jian ZHANG1(), Jianbo YU1, Jian TANG2   

  1. 1.School of Computer Science/ School of Cyber Science and Engineering,Nanjing University of Information Science and Technology,Nanjing Jiangsu 210044,China
    2.School of Information Science and Technology,Beijing University of Technology,Beijing 100124,China
  • Received:2025-04-07 Revised:2025-06-16 Accepted:2025-06-25 Online:2025-07-15 Published:2026-03-10
  • Contact: Jian ZHANG
  • About author:YU Jianbo, born in 2001, M. S. candidate. His research interests include machine learning, image classification.
    TANG Jian, born in 1974, Ph. D., professor. His research interests include small sample data modeling, intelligent control of municipal solid waste treatment process.
  • Supported by:
    National Natural Science Foundation of China(62073006)

摘要:

由于国内城市固废焚烧(MSWI)过程的火焰图像具有强污染、高噪声和曝光度过高等问题,传统目标识别方法难以适用。因此,提出一种MSWI焚烧图像分类框架——基于多层预处理的SAswin网络(SAswin-MPNet)。首先,设计基于Transformer的混合注意力超分辨率重建(HASRT)模块对图像进行超分辨率的重建处理;其次,引入实用曝光校正(PEC)模块对高分辨率MSWI图像进行曝光度校正,从而获得多层预处理数据;此外,设计检验算法将预处理后的图像与原图像进行对照检验,将达到检验阈值的图像替换原图像,从而获得多层预处理数据集;最后,构建SAswin分类网络以识别燃烧状态。基于某MSWI电厂实际运行数据与ResNet-34、ResNet-50、ConvNeXt、ViT(Vision Transformer)、Swin-T(Swin-Tiny)和EVA-02(Enhanced Visual Assistant-02)进行对比的实验结果表明,SAswin-MPNet的MSWI图像燃烧状态识别的准确率与F1分数取得了最优结果。

关键词: 城市固废焚烧, 图像预处理, 图像分类, Transformer, 深度学习

Abstract:

Due to the strong contamination, high noise level, excessive exposure, and other problems in flame images from domestic Municipal Solid Waste Incineration (MSWI) processes, traditional target recognition methods are difficult to apply to them. Therefore, an MSWI incineration image classification framework — SAswin with Multilayer Preprocessing Network (SAswin-MPNet) was proposed. Firstly, a Transformer-based Hybrid Attention Super-Resolution Transformer (HASRT) module was designed to perform super-resolution reconstruction to the images. Secondly, a Practical Exposure Correction (PEC) module was introduced to correct the exposure of high-resolution MSWI images, thereby obtaining multilayer preprocessed data. Additionally, a validation algorithm was designed to compare and test the preprocessed images and the originals, and the images meeting a validation threshold were used to replace the originals, thereby obtaining a multilayer preprocessed dataset. Finally, an SAswin classification network was constructed to recognize incineration states. Experimental results based on actual operational data from an MSWI power plant comparing with ResNet-34, ResNet-50, ConvNeXt, ViT (Vision Transformer), Swin-T (Swin-Tiny), and EVA-02 (Enhanced Visual Assistant-02) show that SAswin-MPNet achieves the optimal MSWI image incineration state recognition accuracy and F1-score.

Key words: Municipal Solid Waste Incineration (MSWI), image preprocessing, image classification, Transformer, deep learning

中图分类号: