Cattle dung detection in pastures from drone images using YOLOv5

IF 1.1 4区农林科学 Q3 AGRICULTURE, MULTIDISCIPLINARY Grassland Science Pub Date : 2024-09-15 DOI:10.1111/grs.12429

Kensuke Kawamura, Yura Kato, Taisuke Yasuda, Eriko Aozasa, Masato Yayota, Miya Kitagawa, Kyoko Kunishige

{"title":"Cattle dung detection in pastures from drone images using YOLOv5","authors":"Kensuke Kawamura, Yura Kato, Taisuke Yasuda, Eriko Aozasa, Masato Yayota, Miya Kitagawa, Kyoko Kunishige","doi":"10.1111/grs.12429","DOIUrl":null,"url":null,"abstract":"<p>Livestock excretions are crucial for nutrient cycling in pasture ecosystems. However, conventional methods based on field observations require significant human power and are time-consuming. This study developed a model, ‘Dung Detector (DD)’, for detecting cattle dung in pastures from drone images using the You Only Look Once (YOLO) v5 algorithm. The DD model was trained using our custom dataset including 1,504 split images from drone orthomosaic images in five paddocks: Obihiro (OBH), Shintoku (STK), Minokamo (MNO), Miyota (MYT), and Yatsugatake (YGK). The detection accuracy was evaluated using ground-truth data acquired in two quadrats within paddocks. The DD model performed well for OBH and STK (<i>F</i>-score = 0.861 and 0.835) paddocks with simple grass species and low surface sward height (SSH). Although the MNO and MYT, with complex vegetation and high SSH, showed few false positives (precision >0.9), some cattle dung pats were undetectable, presumably due to grass height (Recall = 0.500 and 0.276).</p>","PeriodicalId":56078,"journal":{"name":"Grassland Science","volume":"70 4","pages":"168-174"},"PeriodicalIF":1.1000,"publicationDate":"2024-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Grassland Science","FirstCategoryId":"97","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1111/grs.12429","RegionNum":4,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"AGRICULTURE, MULTIDISCIPLINARY","Score":null,"Total":0}

引用次数: 0

Abstract

Livestock excretions are crucial for nutrient cycling in pasture ecosystems. However, conventional methods based on field observations require significant human power and are time-consuming. This study developed a model, ‘Dung Detector (DD)’, for detecting cattle dung in pastures from drone images using the You Only Look Once (YOLO) v5 algorithm. The DD model was trained using our custom dataset including 1,504 split images from drone orthomosaic images in five paddocks: Obihiro (OBH), Shintoku (STK), Minokamo (MNO), Miyota (MYT), and Yatsugatake (YGK). The detection accuracy was evaluated using ground-truth data acquired in two quadrats within paddocks. The DD model performed well for OBH and STK (F-score = 0.861 and 0.835) paddocks with simple grass species and low surface sward height (SSH). Although the MNO and MYT, with complex vegetation and high SSH, showed few false positives (precision >0.9), some cattle dung pats were undetectable, presumably due to grass height (Recall = 0.500 and 0.276).

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

利用 YOLOv5 从无人机图像中检测牧场中的牛粪

牲畜排泄物对牧场生态系统的养分循环至关重要。然而，基于实地观察的传统方法需要大量人力且耗时。本研究开发了一个 "牛粪检测器（DD）"模型，利用YOLO v5算法从无人机图像中检测牧场中的牛粪。DD 模型是利用我们的定制数据集进行训练的，该数据集包括来自五个围场的无人机正射影像的 1,504 张分割图像：带广 (OBH)、新德 (STK)、美浓加茂 (MNO)、宫田 (MYT) 和八岳 (YGK)。利用在围场内两个四分区获取的地面实况数据对检测精度进行了评估。DD 模型在 OBH 和 STK（F-score = 0.861 和 0.835）围场中表现良好，这些围场的草种简单，表面草丛高度（SSH）较低。虽然植被复杂、SSH 高的 MNO 和 MYT 模型很少出现误报（精确度为 0.9），但有些牛粪斑无法检测到，这可能是由于草高造成的（Recall = 0.500 和 0.276）。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Grassland Science Agricultural and Biological Sciences-Agronomy and Crop Science

CiteScore

2.70

自引率

7.70%

发文量

审稿时长

>12 weeks

期刊介绍： Grassland Science is the official English language journal of the Japanese Society of Grassland Science. It publishes original research papers, review articles and short reports in all aspects of grassland science, with an aim of presenting and sharing knowledge, ideas and philosophies on better management and use of grasslands, forage crops and turf plants for both agricultural and non-agricultural purposes across the world. Contributions from anyone, non-members as well as members, are welcome in any of the following fields: grassland environment, landscape, ecology and systems analysis; pasture and lawn establishment, management and cultivation; grassland utilization, animal management, behavior, nutrition and production; forage conservation, processing, storage, utilization and nutritive value; physiology, morphology, pathology and entomology of plants; breeding and genetics; physicochemical property of soil, soil animals and microorganisms and plant nutrition; economics in grassland systems.

期刊最新文献

Issue Information Cattle dung detection in pastures from drone images using YOLOv5 Potassium fertilization and defoliation intensity effects on forage characteristics of “BRS Zuri” guineagrass Phylogenomic identification and overexpression of plant size–related genes in Setaria viridis and rice Issue Information