CGBNet: A Deep Learning Framework for Compost Classification
Minimizing waste entering US landfills has become increasingly difficult. Composting organic waste can assist in reducing the effects of landfill pollution and greenhouse gas emissions. To help automate this process, we apply computer vision techniques to train a classifier to distinguish between gr...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2022-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9864578/ |
_version_ | 1811182915102441472 |
---|---|
author | Suchisrit Gangopadhyay Anthony Zhai |
author_facet | Suchisrit Gangopadhyay Anthony Zhai |
author_sort | Suchisrit Gangopadhyay |
collection | DOAJ |
description | Minimizing waste entering US landfills has become increasingly difficult. Composting organic waste can assist in reducing the effects of landfill pollution and greenhouse gas emissions. To help automate this process, we apply computer vision techniques to train a classifier to distinguish between green and brown compost. The lack of labeled compost data and the challenges and dangers posed by collecting such data create a problem for the automation of composting. In this study, we propose CGBNet, a deep learning framework to differentiate between green and brown compost for the creation of healthy compost piles with an effective ratio of carbon to nitrogen. First, superclass and subclass classification are proposed as two methods of labeling and classifying data. Six deep learning models are then trained and evaluated on a newly formed dataset of 1,960 images using both label structures. Next, transfer learning is applied for feature extraction to boost performance using state-of-the-art deep learning models. The models showed improvements in performance generalization with less overfitting. The best-performing model, CGBNet, used transfer learning and subclass classification to obtain an accuracy of 95%, a 16% improvement in accuracy from the best performing superclass classification model trained without transfer learning. Superclass classification also showed significant improvements in performance and practicality with transfer learning, allowing for robust compost recognition systems to be more extensible while requiring less data and fewer parameters to train. |
first_indexed | 2024-04-11T09:38:26Z |
format | Article |
id | doaj.art-1a87e8399089415fadd52f4db556b839 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-04-11T09:38:26Z |
publishDate | 2022-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-1a87e8399089415fadd52f4db556b8392022-12-22T04:31:21ZengIEEEIEEE Access2169-35362022-01-0110900689007810.1109/ACCESS.2022.32010999864578CGBNet: A Deep Learning Framework for Compost ClassificationSuchisrit Gangopadhyay0https://orcid.org/0000-0001-5685-5014Anthony Zhai1The Hun School of Princeton, Princeton, NJ, USAMontgomery High School, Skillman, NJ, USAMinimizing waste entering US landfills has become increasingly difficult. Composting organic waste can assist in reducing the effects of landfill pollution and greenhouse gas emissions. To help automate this process, we apply computer vision techniques to train a classifier to distinguish between green and brown compost. The lack of labeled compost data and the challenges and dangers posed by collecting such data create a problem for the automation of composting. In this study, we propose CGBNet, a deep learning framework to differentiate between green and brown compost for the creation of healthy compost piles with an effective ratio of carbon to nitrogen. First, superclass and subclass classification are proposed as two methods of labeling and classifying data. Six deep learning models are then trained and evaluated on a newly formed dataset of 1,960 images using both label structures. Next, transfer learning is applied for feature extraction to boost performance using state-of-the-art deep learning models. The models showed improvements in performance generalization with less overfitting. The best-performing model, CGBNet, used transfer learning and subclass classification to obtain an accuracy of 95%, a 16% improvement in accuracy from the best performing superclass classification model trained without transfer learning. Superclass classification also showed significant improvements in performance and practicality with transfer learning, allowing for robust compost recognition systems to be more extensible while requiring less data and fewer parameters to train.https://ieeexplore.ieee.org/document/9864578/Machine learningcompost classificationsuperclass classificationsubclass classificationtransfer learningconvolutional neural network |
spellingShingle | Suchisrit Gangopadhyay Anthony Zhai CGBNet: A Deep Learning Framework for Compost Classification IEEE Access Machine learning compost classification superclass classification subclass classification transfer learning convolutional neural network |
title | CGBNet: A Deep Learning Framework for Compost Classification |
title_full | CGBNet: A Deep Learning Framework for Compost Classification |
title_fullStr | CGBNet: A Deep Learning Framework for Compost Classification |
title_full_unstemmed | CGBNet: A Deep Learning Framework for Compost Classification |
title_short | CGBNet: A Deep Learning Framework for Compost Classification |
title_sort | cgbnet a deep learning framework for compost classification |
topic | Machine learning compost classification superclass classification subclass classification transfer learning convolutional neural network |
url | https://ieeexplore.ieee.org/document/9864578/ |
work_keys_str_mv | AT suchisritgangopadhyay cgbnetadeeplearningframeworkforcompostclassification AT anthonyzhai cgbnetadeeplearningframeworkforcompostclassification |