Jonathan Cox, Xiaodong Li, Charles Fox, Shaun Coutts
{"title":"Black-grass (Alopecurus myosuroides) in cereal multispectral detection by UAV","authors":"Jonathan Cox, Xiaodong Li, Charles Fox, Shaun Coutts","doi":"10.1017/wsc.2023.41","DOIUrl":null,"url":null,"abstract":"\n Site-specific weed management (on the scale of a few meters or less) has the potential to greatly reduce pesticide use and its associated environmental and economic costs. A prerequisite for site-specific weed management is the availability of accurate maps of the weed population that can be generated quickly and cheaply. Improvements and cost reductions in unmanned aerial vehicles (UAVs) and camera technology mean these tools are now readily available for agricultural use. We used UAVs to collect aerial images captured in both RGB and multispectral formats of twelve cereal fields (wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.)) across eastern England. This data was used to train machine learning models to generate prediction maps of black-grass (Alopecurus myosuroides Huds.) locations, a prolific weed in UK cereal fields. We tested machine learning and data set resampling methods to obtain the most accurate system for predicting the presence and absence of weeds in new out-of-sample fields. The accuracy of the system in predicting the absence of A. myosuroides is 69% and its presence above 5g in weight with 77% accuracy in new out-of-sample fields. This system generates prediction maps that can be used by either agricultural machinery or autonomous robotic platforms for precision weed management. Improvements to the accuracy can be made by increasing the number of fields and samples in the data set and the length of time over which data is collected to gather data over the whole growing season.","PeriodicalId":23688,"journal":{"name":"Weed Science","volume":" ","pages":""},"PeriodicalIF":2.1000,"publicationDate":"2023-07-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Weed Science","FirstCategoryId":"97","ListUrlMain":"https://doi.org/10.1017/wsc.2023.41","RegionNum":2,"RegionCategory":"农林科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"AGRONOMY","Score":null,"Total":0}
引用次数: 0
Abstract
Site-specific weed management (on the scale of a few meters or less) has the potential to greatly reduce pesticide use and its associated environmental and economic costs. A prerequisite for site-specific weed management is the availability of accurate maps of the weed population that can be generated quickly and cheaply. Improvements and cost reductions in unmanned aerial vehicles (UAVs) and camera technology mean these tools are now readily available for agricultural use. We used UAVs to collect aerial images captured in both RGB and multispectral formats of twelve cereal fields (wheat (Triticum aestivum L.) and barley (Hordeum vulgare L.)) across eastern England. This data was used to train machine learning models to generate prediction maps of black-grass (Alopecurus myosuroides Huds.) locations, a prolific weed in UK cereal fields. We tested machine learning and data set resampling methods to obtain the most accurate system for predicting the presence and absence of weeds in new out-of-sample fields. The accuracy of the system in predicting the absence of A. myosuroides is 69% and its presence above 5g in weight with 77% accuracy in new out-of-sample fields. This system generates prediction maps that can be used by either agricultural machinery or autonomous robotic platforms for precision weed management. Improvements to the accuracy can be made by increasing the number of fields and samples in the data set and the length of time over which data is collected to gather data over the whole growing season.
期刊介绍:
Weed Science publishes original research and scholarship in the form of peer-reviewed articles focused on fundamental research directly related to all aspects of weed science in agricultural systems. Topics for Weed Science include:
- the biology and ecology of weeds in agricultural, forestry, aquatic, turf, recreational, rights-of-way and other settings, genetics of weeds
- herbicide resistance, chemistry, biochemistry, physiology and molecular action of herbicides and plant growth regulators used to manage undesirable vegetation
- ecology of cropping and other agricultural systems as they relate to weed management
- biological and ecological aspects of weed control tools including biological agents, and herbicide resistant crops
- effect of weed management on soil, air and water.