EPIC: An automated diagnostic tool for Potato Late Blight disease detection from images

EPIC：一种从图像检测马铃薯晚疫病的自动化诊断工具

基本信息

批准号：
BB/R019983/1
负责人：
Liangxiu Han
金额：
$ 8.23万
依托单位：
Manchester Metropolitan University
依托单位国家：
英国
项目类别：
Research Grant
财政年份：
2018
资助国家：
英国
起止时间：
2018 至无数据
项目状态：
已结题

来源：
https://gtr.ukri.org/projects?ref=BB%2FR019983%2F1
关键词：
EPIC automated diagnostic tool Potato

项目摘要

The yields of crop plants are deleteriously affected by various diseases. It is estimated that almost 25% of worldwide crops are lost to diseases, which may cause devastating economical, social and ecological losses. In China, as the fourth important food crop, yield losses from potato late blight diseases can vary from 20%-40% in common years. In severe cases, the yield loss may reach 50%-100%. The estimated yearly economic losses due to this disease are around $5 billion in China. Early accurate detection and identification of crop diseases plays an important role in effectively controlling and preventing diseases for sustainable agriculture and food security.In our previous funded projects, we have developed an innovative automated machine vision system for efficient crop disease diagnosis from images, which have proven the technical feasibility of using advanced image processing, machine learning, mobile and cloud computing approaches. This project will take it forward and develop a near-market product ready for commercialisation, which can provide more accurate real-time information for crop disease surveillance. The tool can run on mobile devices. Farmers with basic training can perform disease diagnosis immediately. Compared to the current practices using human visual observation (which is labour intensive, costly and error-prone), this machine vision system can dramatically speed up diagnosis, and give growers more accurate information on which to base their disease control strategies and stop crop yields from being reduced by infection. This technology can overcome lack of expertise, help make a significant impact on agricultural productivity and farmer incomes, ensuring food security, and deliver highly cost-effective, long-term economic and social impact in China.To achieve actual impact and demonstrable benefits in China, this project will work closely with Chinese partners from academia, industry and farmers including: the project partner (Hebei Agriculture University (HEBAU)), and end users (Beijing Mengbangda Biotechnology Co. Ltd (BMB) and Guyuan County Potato Association (GCPA)). They will provide support in gathering the field data, setting up trial systems and domain knowledge input from plant pathologists for local agriculture and fine tuning the systems in the fields, as well as potential commercialization of this technology in China (Hebei province initially). The project focuses on three stages of translational/user engagement: 1) System requirement gathering from users; 2) System evaluation with input from users; 3) Potential impact and commercial exploitation with end users. The tool will be initially deployed in the real fields provided by end users (BMB and GCPA) in Hebei province at the end of project. This will help protect potato-planting area of over 380k MU initially from the disease infection with reduced annual costs on fungicide usage and damages to environment. Other translational activities include organization of workshops, conference attendance and paper publications, which will be used for dissemination and engagement with a wide range of user groups on a large-scale.This project will not only develop a near-market product but will also generate a significantly measurable impact, promote long-term sustainable growth, economic development and welfare in China and beyond.

作物植物的产量受到各种疾病的有害影响。据估计，几乎25％的全球作物因疾病而丢失，这可能会导致经济，社会和生态损失。在中国，作为第四种重要的粮食作物，马铃薯晚疫病疾病的产量损失可能从普通年份的20％-40％不等。在严重的情况下，收益率损失可能达到50％-100％。在中国，由于这种疾病造成的估计每年经济损失约为50亿美元。早期准确的检测和鉴定作物疾病在有效控制和预防可持续农业和粮食安全的疾病中起着重要的作用。在我们以前的资助项目中，我们开发了一种创新的自动化机器视觉系统，可从图像中有效地从图像中诊断出作物疾病，这些系统证明了使用高级图像处理，机器学习，机器学习，移动和云计算方法的技术可行性。该项目将把它推向前进，并开发了准备商业化的近市场产品，该产品可以为作物疾病监测提供更准确的实时信息。该工具可以在移动设备上运行。接受基础训练的农民可以立即进行疾病诊断。与使用人类视觉观察（劳动力密集，昂贵且容易出错）的当前做法相比，这种机器视觉系统可以大大加快诊断的速度，并为种植者提供更准确的信息，以基于其疾病控制策略和停止作物产量的基础，并通过感染降低作物。 This technology can overcome lack of expertise, help make a significant impact on agricultural productivity and farmer incomes, ensuring food security, and deliver highly cost-effective, long-term economic and social impact in China.To achieve actual impact and demonstrable benefits in China, this project will work closely with Chinese partners from academia, industry and farmers including: the project partner (Hebei Agriculture University (HEBAU)), and end users (Beijing Mengbangda Biotechnology Co. Ltd（BMB）和Guyuan County Mota Association（GCPA））。他们将提供支持，以收集现场数据，从植物病理学家那里设置试验系统和域知识输入，以便在当地农业和对现场中的系统进行微调，以及该技术在中国（最初是Hebei Province）的潜在商业化。该项目重点介绍了转化/用户参与度的三个阶段：1）从用户收集的系统需求； 2）系统评估用户的输入； 3）对最终用户的潜在影响和商业剥削。该工具最初将部署在项目结束时在Hebei省最终用户（BMB和GCPA）提供的实际领域中。这将有助于保护超过380k MU的马铃薯植入区域免受疾病感染的影响，而杀菌剂的使用情况下的年成本降低和对环境的损害。其他翻译活动包括组织研讨会，会议出勤和纸质出版物，这些活动将用于大规模上的广泛用户群体的传播和互动。该项目不仅将开发近市场产品，而且还将产生可观的可测量影响，促进长期的可持续性增长，中国和超越的经济发展，经济发展和福利。

项目成果

期刊论文数量（10）

专著数量（0）

科研奖励数量（0）

会议论文数量（0）

专利数量（0）

A Biologically Interpretable Two-Stage Deep Neural Network (BIT-DNN) for Vegetation Recognition From Hyperspectral Imagery

DOI：
10.1109/tgrs.2021.3058782
发表时间：
2021-02
期刊：
IEEE Transactions on Geoscience and Remote Sensing
影响因子：
8.2
作者：
Yue Shi;Liangxiu Han;Wenjiang Huang;Sheng Chang;Yingying Dong;D. Dancey;Lianghao Han
通讯作者：
Yue Shi;Liangxiu Han;Wenjiang Huang;Sheng Chang;Yingying Dong;D. Dancey;Lianghao Han

Novel CropdocNet Model for Automated Potato Late Blight Disease Detection from Unmanned Aerial Vehicle-Based Hyperspectral Imagery