A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision Agriculture

In this paper, we propose a knowledge distillation framework specifically designed for semantic segmentation tasks in agricultural scenarios. This framework aims to address several prevalent challenges in smart agriculture, including limited computational resources, strict real-time constraints, and...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhiyong Li, Lan Xiang, Jun Sun, Dingyi Liao, Lijia Xu, Mantao Wang
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Agriculture
Subjects:
Online Access:https://www.mdpi.com/2077-0472/15/13/1418
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850118416108617728
author Zhiyong Li
Lan Xiang
Jun Sun
Dingyi Liao
Lijia Xu
Mantao Wang
author_facet Zhiyong Li
Lan Xiang
Jun Sun
Dingyi Liao
Lijia Xu
Mantao Wang
author_sort Zhiyong Li
collection DOAJ
description In this paper, we propose a knowledge distillation framework specifically designed for semantic segmentation tasks in agricultural scenarios. This framework aims to address several prevalent challenges in smart agriculture, including limited computational resources, strict real-time constraints, and suboptimal segmentation accuracy on cropped images. Traditional single-level feature distillation methods often suffer from insufficient knowledge transfer and inefficient utilization of multi-scale features, which significantly limits their ability to accurately segment complex crop structures in dynamic field environments. To overcome these issues, we propose a multi-level distillation strategy that leverages feature and embedding patch distillation, combining high-level semantic features with low-level texture details for joint distillation. This approach enables the precise capture of fine-grained agricultural elements, such as crop boundaries, stems, petioles, and weed clusters, which are critical for achieving robust segmentation. Additionally, we integrated an enhanced attention mechanism into the framework, which effectively strengthens and fuses key crop-related features during the distillation process, thereby further improving the model’s performance and image understanding capabilities. Extensive experiments on two agricultural datasets (sweet pepper and sugar) demonstrate that our method improves segmentation accuracy by 7.59% and 6.79%, without significantly increasing model complexity. Further validation shows that our approach exhibits strong generalization capabilities on two widely used public datasets, proving its applicability beyond agricultural domains.
format Article
id doaj-art-60aef9bc67ab4d2faf1ca56e5b3607f0
institution OA Journals
issn 2077-0472
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Agriculture
spelling doaj-art-60aef9bc67ab4d2faf1ca56e5b3607f02025-08-20T02:35:53ZengMDPI AGAgriculture2077-04722025-06-011513141810.3390/agriculture15131418A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision AgricultureZhiyong Li0Lan Xiang1Jun Sun2Dingyi Liao3Lijia Xu4Mantao Wang5College of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaObservation and Research Station of Land Ecology and Land Use in Chengdu Plain, Ministry of Natural Resources, Chengdu 610045, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaCollege of Mechanical and Electrical Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaCollege of Information Engineering, Sichuan Agricultural University, Ya’an 625000, ChinaIn this paper, we propose a knowledge distillation framework specifically designed for semantic segmentation tasks in agricultural scenarios. This framework aims to address several prevalent challenges in smart agriculture, including limited computational resources, strict real-time constraints, and suboptimal segmentation accuracy on cropped images. Traditional single-level feature distillation methods often suffer from insufficient knowledge transfer and inefficient utilization of multi-scale features, which significantly limits their ability to accurately segment complex crop structures in dynamic field environments. To overcome these issues, we propose a multi-level distillation strategy that leverages feature and embedding patch distillation, combining high-level semantic features with low-level texture details for joint distillation. This approach enables the precise capture of fine-grained agricultural elements, such as crop boundaries, stems, petioles, and weed clusters, which are critical for achieving robust segmentation. Additionally, we integrated an enhanced attention mechanism into the framework, which effectively strengthens and fuses key crop-related features during the distillation process, thereby further improving the model’s performance and image understanding capabilities. Extensive experiments on two agricultural datasets (sweet pepper and sugar) demonstrate that our method improves segmentation accuracy by 7.59% and 6.79%, without significantly increasing model complexity. Further validation shows that our approach exhibits strong generalization capabilities on two widely used public datasets, proving its applicability beyond agricultural domains.https://www.mdpi.com/2077-0472/15/13/1418knowledge distillationsemantic segmentationmulti-scale featuresembedding patchattention mechanismgeneralization capability
spellingShingle Zhiyong Li
Lan Xiang
Jun Sun
Dingyi Liao
Lijia Xu
Mantao Wang
A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision Agriculture
Agriculture
knowledge distillation
semantic segmentation
multi-scale features
embedding patch
attention mechanism
generalization capability
title A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision Agriculture
title_full A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision Agriculture
title_fullStr A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision Agriculture
title_full_unstemmed A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision Agriculture
title_short A Multi-Level Knowledge Distillation for Enhanced Crop Segmentation in Precision Agriculture
title_sort multi level knowledge distillation for enhanced crop segmentation in precision agriculture
topic knowledge distillation
semantic segmentation
multi-scale features
embedding patch
attention mechanism
generalization capability
url https://www.mdpi.com/2077-0472/15/13/1418
work_keys_str_mv AT zhiyongli amultilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT lanxiang amultilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT junsun amultilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT dingyiliao amultilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT lijiaxu amultilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT mantaowang amultilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT zhiyongli multilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT lanxiang multilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT junsun multilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT dingyiliao multilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT lijiaxu multilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture
AT mantaowang multilevelknowledgedistillationforenhancedcropsegmentationinprecisionagriculture