ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea Detection

During the harvesting and preparation of tea, it is common for tea to become mixed with some impurities. Eliminating these impurities is essential to improve the quality of famous green tea. At present, this sorting procedure heavily depends on manual efforts, which include high labor intensity, low...

Full description

Saved in:
Bibliographic Details
Main Authors: Zezhong Ding, Yanfang Li, Bin Hu, Zhiwei Chen, Houzhen Jia, Yali Shi, Xingmin Zhang, Xuesong Zhu, Wenjie Feng, Chunwang Dong
Format: Article
Language:English
Published: MDPI AG 2025-04-01
Series:Foods
Subjects:
Online Access:https://www.mdpi.com/2304-8158/14/9/1554
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850137646132625408
author Zezhong Ding
Yanfang Li
Bin Hu
Zhiwei Chen
Houzhen Jia
Yali Shi
Xingmin Zhang
Xuesong Zhu
Wenjie Feng
Chunwang Dong
author_facet Zezhong Ding
Yanfang Li
Bin Hu
Zhiwei Chen
Houzhen Jia
Yali Shi
Xingmin Zhang
Xuesong Zhu
Wenjie Feng
Chunwang Dong
author_sort Zezhong Ding
collection DOAJ
description During the harvesting and preparation of tea, it is common for tea to become mixed with some impurities. Eliminating these impurities is essential to improve the quality of famous green tea. At present, this sorting procedure heavily depends on manual efforts, which include high labor intensity, low sorting efficiency, and high sorting costs. In addition, the hardware performance is poor in actual production, and the model is not suitable for deployment. To solve this technical problem in the industry, this article proposes a lightweight algorithm for detecting and sorting impurities in premium green tea in order to improve sorting efficiency and reduce labor intensity. A custom dataset containing four categories of impurities was created. This dataset was employed to evaluate various YOLOv8 models, ultimately leading to the selection of YOLOv8n as the base model. Initially, four loss functions were compared in the experiment, and Focaler_mpdiou was chosen as the final loss function. Subsequently, this loss function was applied to other YOLOv8 models, leading to the selection of YOLOv8m-Focaler_mpdiou as the teacher model. The model was then pruned to achieve a lightweight model at the expense of detection accuracy. Finally, knowledge distillation was applied to enhance its detection performance. Compared to the base model, it showed advancements in P, R, mAP, and FPS by margins of 0.0051, 0.0120, and 0.0094 and an increase of 72.2 FPS, respectively. Simultaneously, it achieved a reduction in computational complexity with GFLOPs decreasing by 2.3 and parameters shrinking by 860350 B. Afterwards, we further demonstrated the model’s generalization ability in black tea samples. This research contributes to the technological foundation for sophisticated impurity classification in tea.
format Article
id doaj-art-d15b452d449e4eed9d2169fc4d413027
institution OA Journals
issn 2304-8158
language English
publishDate 2025-04-01
publisher MDPI AG
record_format Article
series Foods
spelling doaj-art-d15b452d449e4eed9d2169fc4d4130272025-08-20T02:30:46ZengMDPI AGFoods2304-81582025-04-01149155410.3390/foods14091554ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea DetectionZezhong Ding0Yanfang Li1Bin Hu2Zhiwei Chen3Houzhen Jia4Yali Shi5Xingmin Zhang6Xuesong Zhu7Wenjie Feng8Chunwang Dong9Tea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaWeifang Engineering Vocational College, Weifang 262500, ChinaCollege of Mechanical and Electronic Engineering, Shihezi University, Shihezi 832000, ChinaTea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaTea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaTea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaTea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaTea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaTea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaTea Research Institute, Information and Economy Institution, Shandong Academy of Agricultural Sciences, Jinan 250100, ChinaDuring the harvesting and preparation of tea, it is common for tea to become mixed with some impurities. Eliminating these impurities is essential to improve the quality of famous green tea. At present, this sorting procedure heavily depends on manual efforts, which include high labor intensity, low sorting efficiency, and high sorting costs. In addition, the hardware performance is poor in actual production, and the model is not suitable for deployment. To solve this technical problem in the industry, this article proposes a lightweight algorithm for detecting and sorting impurities in premium green tea in order to improve sorting efficiency and reduce labor intensity. A custom dataset containing four categories of impurities was created. This dataset was employed to evaluate various YOLOv8 models, ultimately leading to the selection of YOLOv8n as the base model. Initially, four loss functions were compared in the experiment, and Focaler_mpdiou was chosen as the final loss function. Subsequently, this loss function was applied to other YOLOv8 models, leading to the selection of YOLOv8m-Focaler_mpdiou as the teacher model. The model was then pruned to achieve a lightweight model at the expense of detection accuracy. Finally, knowledge distillation was applied to enhance its detection performance. Compared to the base model, it showed advancements in P, R, mAP, and FPS by margins of 0.0051, 0.0120, and 0.0094 and an increase of 72.2 FPS, respectively. Simultaneously, it achieved a reduction in computational complexity with GFLOPs decreasing by 2.3 and parameters shrinking by 860350 B. Afterwards, we further demonstrated the model’s generalization ability in black tea samples. This research contributes to the technological foundation for sophisticated impurity classification in tea.https://www.mdpi.com/2304-8158/14/9/1554teaimpurity identificationdeep learningFocaler_mpdioumodel pruningknowledge distillation
spellingShingle Zezhong Ding
Yanfang Li
Bin Hu
Zhiwei Chen
Houzhen Jia
Yali Shi
Xingmin Zhang
Xuesong Zhu
Wenjie Feng
Chunwang Dong
ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea Detection
Foods
tea
impurity identification
deep learning
Focaler_mpdiou
model pruning
knowledge distillation
title ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea Detection
title_full ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea Detection
title_fullStr ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea Detection
title_full_unstemmed ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea Detection
title_short ITD-YOLO: An Improved YOLO Model for Impurities in Premium Green Tea Detection
title_sort itd yolo an improved yolo model for impurities in premium green tea detection
topic tea
impurity identification
deep learning
Focaler_mpdiou
model pruning
knowledge distillation
url https://www.mdpi.com/2304-8158/14/9/1554
work_keys_str_mv AT zezhongding itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT yanfangli itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT binhu itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT zhiweichen itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT houzhenjia itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT yalishi itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT xingminzhang itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT xuesongzhu itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT wenjiefeng itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection
AT chunwangdong itdyoloanimprovedyolomodelforimpuritiesinpremiumgreenteadetection