Produced by the International Association for the Plant Protection Sciences (IAPPS). To join IAPPS and receive the Crop Protection journal online go to: www.plantprotection.org
ITP collaborator: Washington State Department of Agriculture
Authors: Quinlyn Baine, Chris Looney, Spencer K. Monckton, David R. Smith, Nathan M. Schiff, Henri Goulet, and Amanda J. Redford
USDA Animal and Plant Health Inspection Service’s Identification Technology Program (ITP) is pleased to announce the final release of Sawfly GenUS. The first release (February 2020) included identification support for all but the most diverse families of sawflies; this release adds that family and a new searchable host plant list. The website includes fact sheets, images, and keys to support a wide variety of users.
Please find the attached PDF announcement to see an overview of ITP’s newest identification tool for PPQ and its partners. Please also feel free to forward this email or the attachment to your colleagues.
Visit our website to learn more about ITP’s tools and mobile apps
Interested in assisting ITP with tool development by being a beta reviewer for an upcoming ITP tool? We are seeking to increase our pool of beta reviewers for a variety of pest groups. Beta reviewers can be experts or non-specialists. Please contact us at itp@usda.gov. If you did not receive this email directly from ITP, and you would like to be included in future ITP announcement emails, please send a request to itp@usda.gov
The visual and tactile examination of plant leaves is a standard method for identifying disease in crops and horticultural products. However, such an approach can be highly subjective and is dependent on the skills of the examiners. Writing in the International Journal of Computational Vision and Robotics, a team from Egypt describes a new approach to plant leaf disease detection using deep learning on a mobile device. The team’s tests against a standard database of diseased leaf images showed their system to be capable of up to 98% diagnostic accuracy. The process is rapid and showcases the sophisticated computational power available in modern mobile phones for this kind of intensive task.
Shaheera A. Rashwan and Marwa K. Elteir of the Informatics Research Institute at the City of Scientific Research and Technological Applications in Alexandria, suggest that for busy farmers in remote regions with no immediate access to plant disease experts, a mobile application that can help them spot disease and so treat the crops in a timely manner could be vital to their ongoing agricultural viability.
The team’s approach exploits the recent evolution of computational systems and especially graphical processing units (GPUs) that allow machine learning operations to be carried out efficiently in ways that previous generations of devices simply could not match for speed. Such operations facilitate the running of tools such as convolutional neural networks, which mimic certain characteristics of brain function, and allow image recognition and related tasks to be carried out quickly. The team thus embedded image recognition of the characteristics of disease in leaves for the present research.
Despite the great speed and accuracy of disease diagnostics that the team has shown, there is still room for improvement. They highlight an issue with shadows on images and confusing backgrounds when a user takes a photo of a suspect leaf. They hope to be able to develop a pre-processing step that will reduce any problems and the inaccuracies that might arise if the acquired leaf image is not as perfect as it might be for image recognition. Fundamentally, automated light level adjustment in the image would preclude issues arising because of shadows, while a step that isolates the leaf from its background in the image and effectively removes said background would ease the whole process still further and hopefully nudge the accuracy upwards.
More information: Shaheera A. Rashwan et al, Plant leaf disease detection using deep learning on mobile devices, International Journal of Computational Vision and Robotics (2022). DOI: 10.1504/IJCVR.2022.121151
This was written by Esther Ngumbi, and appeared on Sci Dev Net.
USAID recently offered prize money for the best digital tools that can be used to help combat the fall armyworm (FAW), an invasive pest that has spread across Africa. The winners will be announced in the coming months.
Map of areas affected by Fall Armyworm (as of January 2018) Credit: FAO
But this is not the first invasive pest the African continent is dealing with. Just a few years ago, African smallholder farmers battled the invasive South American tomato moth, Tuta absoluta. According to recent research, five invasive insect pests including T. absoluta cost the African continent US$ 1.1 billion every year.
Around the world, invasive pests are causing US$ 540 billion in economic losses to agriculture each year despite the fact that many countries are doing their best to prevent insect invasions now and into the future.
Tackling invasive pests reactively
To deal with invasive insects, African countries assisted by other stakeholders, including aid agencies such as USAID, research institutions such as the International Center for Insect Physiology and Ecology, the Center for Agriculture and Bioscience International (CABI, the parent organization of SciDev.Net) and the United Nations Food and Agriculture Organization (UN FAO) have repeatedly taken a reactive rather than a proactive approach in tackling the invasive pests only after they have established a foothold and caused considerable damage.
Ghana, for example, established a National Taskforce to control and manage FAW after the worms had invaded local fields. This taskforce mandate includes sensitizing farmers and making them aware of the symptoms of armyworm attacks so they can report infestations to authorities and undertake research aimed at finding short and long term solutions to combat the spread of FAW.
“While many of these strategies are working, one cannot help but wonder what it would take for African governments to get ahead of this problem.”
Esther Ngumbi, University of Illinois
Malawi’s government prioritized the use of pesticides as an immediate and short-term strategy to fight the FAW after many of their smallholder farmers lost crops to this invasive insect. Further, the government intensified training and awareness campaigns about this pest and installed pheromone traps to help monitor the spread only after the pest had established a foothold.
The FAO, a leader in the efforts to deal with invasive pests in Africa, has spearheaded many efforts including bringing together experts from the Americas, Africa and other regions to share and update each other on FAW. The FAO has launched a mobile phone app to be used as an early warning system tool. But again, many of these efforts happened after the first detection of the FAW.
While many of these strategies are working, one cannot help but wonder what it would take for African governments to get ahead of this problem. How can aid agencies such as USAID, UN FAO and other development partners that are currently spending billions to fight the invasive FAW help Africa to take the necessary steps to ensure that it is better prepared to deal with invasive insects now and into the future?
Anticipate and prepare
Recent research predicts that threats from invasive insects will continue to increase with African countries expected to be the most vulnerable. African governments must anticipate and prepare for such invasions using already available resources.
Early this year, CABI launched invasive species Horizon Scanning Tool (beta), a tool that allows countries to identify potential invasive species. This online and open source tool supported by United States Department of Agriculture and the UK Department for International Development allows countries to generate a list of invasive species that are absent from their countries at the moment but present in “source areas,” which may be relevant because they are neighboring countries, linked by trade and transport routes, or share similar climates. Doing so could allow African countries to prepare action plans that can be quickly rolled out when potential invaders actually arrive.
Learn from other regions
Africa can learn from other regions that have comprehensive plans on dealing with invasive insects and countries that have gone through similar invasions. The United States and Australia are examples of countries that have comprehensive plans on preventing and dealing with insect invasions, while Brazil has gone through its own FAW invasion.
“African governments must learn to be proactive rather than reactive in dealing with invasive insects.”
Esther Ngumbi, University of Illinois
Through workshops and training programs that help bring experts together, African countries can learn how to prevent and deal with future insect invasions. Moreover, key actors should help organize more workshops and training programs to enable African experts to learn from their counterparts overseas. At the same time, the manuals, and all the information exchanged and learned during such workshops, could be stored in online repositories that can be accessed by all African countries.
Strengthen African pest surveillance
A recent Feed the Future funded technical brief, which I helped to write, looked at the strength of existing African plant protection regulatory frameworks by examining eight indicators including the existence of a specified government agency mandated with the task of carrying out pest surveillance.
It reveals that many African countries have weak plant protection regulatory systems and that many governments do not carry out routine pest surveillance which involves the collection, recording, analysis, interpretation and timely dissemination of information about the presence, prevalence and distribution of pests.
The International Plant Protection Convention offers a comprehensive document that can help African countries to design pest surveillance programs. Also, the convention offers other guiding documents that can be used by African countries to strengthen their plant protection frameworks. African countries can use these available documents to strengthen national and regional pest surveillance abilities.
Set up emergency funds
Invasive insects know no borders. Thus, African countries must work together. At the same time, given the rapid spread of invasive insect outbreaks, the African continent must set up an emergency fund that can easily be tapped when insects invade. In dealing with the recent FAW invasion, it was evident that individual African countries and the continent did not have an emergency financing plan. This must change.
By anticipating potential invasive insects and learning from countries that have comprehensive national plant protection frameworks, Africa can be prepared for the next insect invasion. African governments must learn to be proactive rather than reactive in dealing with invasive insects.
Doing so will help safeguard Africa’s agriculture and protect the meaningful gains made in agricultural development. Time is ripe.
Esther Ngumbi is a distinguished postdoctoral researcher with the Department of Entomology at the US-based University of Illinois at Urbana Champaign, a World Policy Institute Senior Fellow, Aspen Institute New Voices Food Security Fellow and a Clinton Global University Initiative Agriculture Commitments Mentor and Ambassador. She can be contacted at enn0002@tigermail.auburn.edu
This piece was produced by SciDev.Net’s Sub-Saharan Africa English desk.
The plant health information collected through Plantwise plant clinics is a valuable resource. The Department of Plant Protection has set up a new National Data Centre in Karachi to collate clinic data. Plant health partners, stakeholders and other knowledge delivery systems will be able to use this information to make evidence-based decisions that strengthen plant health systems.
Farmer in Pakistan
Techniques for managing data have become essential and are now common in business and services. With this in mind, Plantwise has now established the first-ever National Data Centre in Karachi, for the assimilation of Plantwise data in Pakistan. This robust management system will collate data coming from more than 1,000 plant clinics throughout the country.
National Data Centre training
The Plantwise programme has provided data management training to plant clinic staff in order to ensure the long-term collection of data. These staff members include local implementing organisation such as agriculture extension departments in each of the Pakistan provinces.
As well as consolidating all the data and information, the National Data Centre aims to offer central hosting and managed services for all Government institutions including the Plantwise programme’s partners.
Data-driven decision-making
The new data centre will enable reliable, efficient and secure access to plant clinics’ plant health data, meaning partner organisations will no longer be required to host their own applications. In addition, partners will receive quicker and more secure networking facilities that will allow them to better showcase their services to end customers.
The National Data Centre will demonstrate how data-driven decision-making is essential for increasing production and improving quality in agriculture.
Setting up the National Data Centre, Karachi
The aim of Planwise was to increase food security and improve rural livelihoods by reducing crop losses. This has been achieved by establishing sustainable networks of local plant clinics, run by trained plant doctors, where farmers receive practical plant health advice. Data collected from these clinics allow for better response and monitoring especially in case of new pest outbreaks, and plant health data is made available for decision making while it is still relevant.
Consolidating data – National Data Centre
It was evident that the data collected from the provincial plant clinics in Pakistan needed to be securely stored with proper access controls. Many of the individual government organisations, such as agricultural extension departments, had set up their own servers, or rather data centres, with high human resource and operational costs.
Therefore, it was considered vital that CABI, in consultation with its LIOI’s, establish a National Data Centre to reduce the cost of data management and increase the value of the clinics’ data. The National Data Centre will also provide the opportunity for cross-validation and high data security.
Help inform plant health services
The plant clinic data coming into the National Data Centre can be used to inform plant health services in various ways. For instance, this data can help to identify new and emerging plant health problems and act as an early warning system to the regulatory bodies, such as the Department of Plant Protection, that are responsible for pest and disease surveillance and response.
CABI’s new global programme, PlantwisePlus, will support low and lower-middle-income countries to predict, prepare themselves for and prevent plant health threats in a changing climate – reducing crop losses and empowering farmers to increase income, food security and food safety by producing more and higher quality food.
PlantwisePlus will therefore focus on (a) strengthening detection of and response to pest outbreaks; (b) providing public and private agricultural service providers with better digital advisory tools to support farmers in sustainable crop management; (c) enhancing the availability of nature-positive and low-risk plant protection products to reduce reliance on high-risk farm inputs; (d) increasing consumer demand for and supply to local markets of safer, higher quality and locally produced food.Pakistan, Plantwise, data managementAgriculture and International Development
Crop leaf disease identification based on ensemble classification
Livestock and horticulture are well-known contributors to the global economy, particularly in countries where farming is the sole motivation for income. Yet, it is regretful that infection degeneration has affected this. Vegetables are a significant source of power for people and animals. Leaves and stems are the most common way for plants to interact with the surroundings. As a consequence, researchers and educators are responsible for investigating the problem and developing ways for recognizing disease-infected leaves.
Growers everywhere across the world will be able to take immediate action to avoid their produce from getting heavily affected, so sparing the globe and themselves from a potential global recession. Because manually diagnosing ailments might not have been the ideal solution, a mechanical methodology for recognizing leaf ailments could benefit the agricultural sector while also enhancing crop output. The goal of this research is to evaluate classification outcomes by combining composite classification with hybrid Law’s mask, LBP, and GLCM.
The proposed method illustrates that a group of classifiers can surpass individual classifiers. The attributes employed are also vital in attaining the best findings because ensemble classification has demonstrated to be much more reliable.
The corn earworm (Helicoverpa zea) is an emerging pest of commercial hemp production throughout the U.S. Boring through stalks and feeding on reproductive structures, this pest presents several management challenges for hemp producers. While integrated pest management strategies for more traditional agricultural crops are established, much work is still needed to develop effective IPM for the corn earworm in hemp. Shown here is a corn earworm larva feeding on a hemp plant flower bud. (Photo originally published in Britt et al 2021, Journal of Integrated Pest Management)
By David Coyle, Ph.D.
David Coyle, Ph.D.
Cheech and Chong. The Big Lebowski. Seth Rogen. These Hollywood legends helped thrust Cannabis sativa into modern-day pop culture, making it simultaneously famous and infamous. And while the topic of C. sativa tends to elicit a range of emotions and opinions, there is no debating the fact this plant has many, many attributes and qualities.
Cannabis sativa is an annual herbaceous crop native to east Asia but is now grown worldwide and can be cultivated for a variety of purposes. Cannabis sativa is known colloquially as hemp or marijuana; these are different cultivars of the same species. The difference between hemp and marijuana is purely chemical: marijuana has a high THC (tetrahydrocannabinol, aka the intoxicating part) content, whereas the THC content in hemp, by definition, must be less than 0.3 percent. There are also some physical differences, as marijuana and hemp grown for cannabinoids have more of a bushy, horticultural crop look while hemp grown for grain or fiber appears more like a row crop, growing from a tall singular stalk.
Hemp became a legal crop with the passage of the 2018 Farm Bill (the Farm Bill is now called the Plant Protection Act, or PPA 7721). This action was significant, as it was the first time hemp was legally differentiated from marijuana. While the law placed restrictions on its production and use, the legalization of industrial hemp (as it is known) led to several pilot production programs being initiated.
Since industrial hemp had not been cultivated in the U.S. before, pest management in this new crop was an area in dire need of research. Several well-known pests present challenges to hemp cultivation, including the corn earworm (Helicoverpa zea). An article published in September in the Journal of Integrated Pest Management highlights what we do and don’t know about H. zea management in industrial hemp. I spoke with the lead author, Kadie Britt, Ph.D., postdoctoral scholar at the University of California, Riverside, about challenges and opportunities associated with this well-known corn (and now hemp) pest.
Coyle: Do you think hemp will take off? I mean, do we even have the infrastructure to support this industry?
The corn earworm (Helicoverpa zea) is an emerging pest of commercial hemp production throughout the U.S. Boring through stalks and feeding on reproductive structures, this pest presents several management challenges for hemp producers. While integrated pest management strategies for more traditional agricultural crops are established, much work is still needed to develop effective IPM for the corn earworm in hemp. Shown here is a corn earworm larva that has tunneled into a hemp plant stem. (Photo originally published in Britt et al 2021, Journal of Integrated Pest Management)
Britt: Yes and no. There is a market, but there’s already too much planted acreage, and the market is saturated. 2019 was a big year after legalization in 2018, and some growers still have that material in their barns as of summer 2021. Hemp is very useful, and there’s a very positive long-term future with this crop in terms of grain and fiber. The fiber can be used for many things, including plastics, clothes, rope, and a bunch of other things. The grain can be used as a food (think Whole Foods fancy spices section) and has very beneficial fatty acids. Unfortunately, we still see too much cannabinoid and not enough grain or fiber acreage.
Regarding management, is it fair to say there are more questions than answers at this point?
Yes! Every answer seems to lead to more questions. Can we sample for pests in hemp as we do in other row crops? Yes, sampling is similar, but with nuances. For instance, pheromone traps don’t seem to be effective, so we’ll have to develop something else. The takeaway is that it’s difficult to rely solely on chemical control, and the best thing we can do or recommend at this point is to watch for eggs and larvae and to initiate spray applications with a product legal for use in hemp. Weekly spraying can be effective but targets only corn earworm—a more IPM-friendly approach is needed, but first we need to know more about the system.
What are the biggest management challenges for industrial hemp?
There are so many! Pest management is a huge challenge, but growers need to be able to successfully produce the crop first. Right now, there’s a lack of infrastructure for the crop as a whole. Industrial hemp is a legitimate alternative to many products; anything from fabric to plastic can be made from hemp. Some companies are building processing facilities, machinery, and all the other infrastructure necessary for a new crop. Having properly labeled pesticides available is a challenge, as industrial hemp probably won’t garner the attention of huge chemical companies, but smaller, newer biopesticide companies may be more willing, as will those that focus on specialty crops.
Any final thoughts?
Yes, the industry is new and emerging, and we have to realize that federal legalization of high-THC cannabis will likely happen at one point. The work we do now will only help prepare us for that time. Cannabis is here to stay, and pest susceptibility greatly increases when the crop is grown in a monoculture type of production. Commercial acreages are different than backyard growing operations. Any information we get now will only help future C. sativa growers, regardless of the final product.
David Coyle, Ph.D., is an assistant professor in the Department of Forestry and Environmental Conservation at Clemson University. Twitter/Instagram/TikTok: @drdavecoyle. Email: dcoyle@clemson.edu.
We are excited to tell you that PestNet has joined forces with the Pacific Pests, Pathogens & Weeds app (compiled by PestNet). It seemed sensible to put these two Pestnet endeavours together. Some time ago, we mentioned that the website had been redesigned to reflect the changes; now we have completed the amalgamation with new mobile apps.
You can see the changes if you visit the website here. And you can download the new mobile apps by searching for “PestNet” or “Pacific Pests, Pathogens & Weeds” from the Google and Apple stores.
[Courtesy of Gyeonggido Business & Science Accelerator] SEOUL — Soybean virus diseases result in reduced yields and inferior quality. Polymerase chain reaction (PCR), a method used in molecular biology, has been widely used to diagnose bean viruses, but it takes up to two weeks to acquire results. South Korean researchers have developed an envelope protein screening method that can easily identify bean viruses.
The common mosaic virus is seed-borne and causes mosaic and lesions on foliage as well as blackened roots. The yellow mosaic virus is often associated with the presence of virus source plants, with symptoms consisting of leaf mosaic formed by contrasting yellow or green mosaic areas.
After dropping a sample collected by scratching the skin of a bean into the diagnostic kit, it will be possible to check the results in five minutes. It enabled the selection of mini antibody candidates without a protein purification process, the bio-center of Gyeonggido Business & Science Accelerator (GBSA) said, adding the commercialization of diagnostic devices based on GBSA’s technology would replace imports.
The center said its research team proposed a new platform for the production of plant virus diagnostic kits using the combined module of SpyTag (peptide) and SpyCatcher (protein) for the binding of antigens and coat proteins. Using SpyTag and SpyCatcher, bioconjugation can be achieved between two recombinant proteins that would otherwise be restrictive or impossible with the traditional direct genetic fusion between the two proteins.
“Through continuous research and development, it has become possible to diagnose the virus in the early stages and preemptively block it,” an unnamed GBSA official said in a statement. “We will contribute to stable agricultural production by reducing economic losses caused by crop virus damage.” The study was conducted jointly with researchers from Kyungpook National University, the Rural Development Administration and the Ajou University College of Medicine.
The Plantwise Knowledge Bank brings together plant health information from across the world. It includes a diagnostic tool, factsheet library, pesticide lists and pest alerts. For those seeking to diagnose a pest problem, the Plantwise Knowledge Bank’s Pest Diagnostic Tool is particularly useful, providing information to help identify the symptoms observed on a crop.
To be able to provide effective solutions to a plant health problem, it’s first important to diagnose a pest problem accurately. The diagnostic tool allows you to diagnose a crop problem through the symptoms observed and the part of the plant affected.
From the Plantwise Knowledge Bank home page, navigate to the diagnose a pest problem tile and click ‘identify a pest’.
Navigate to ‘Identify a pest’
Identify the pest problem
Once clicking on ‘identify a pest’, you are able to search by country or region from the drop-down list. The crop name is then typed into the search box and a list of suggested crops will appear to choose from.
Search by country and crop
Narrow down cause
The first step in narrowing down the cause of a plant health problem is to determine which part of the plant is affected by the pest. In some cases, several parts of the plant can be damaged, but to help diagnose the problem, the main part of the affected plant needs to be determined. The options provided in the tool include leaves, stems, whole plant, seeds, fruit, growing point, inflorescence, roots and vegetative organs.
Main part of affected plant selection
If the type of problem is already known it will help to narrow down your search further, otherwise a user would select ‘unsure’. The types of problems in the tool include mites, insects, fungi, nematodes and weeds.
The steps in the diagnostic tool mirror the plant clinic prescription form found on the data collection app.
Type of problem selection
Pest and disease results
Results from the diagnostic search are given as a list of possible pests or diseases, each with an image, and a link to a technical factsheet further describing the problem. The technical factsheet provides information on crop symptoms, preventative methods and effective solutions to the problem.
Given Tanzania’s diverse geographical landscape, it’s no surprise the country is among the world’s top 20 producers of vegetables. Nevertheless, farmers remain in search of ways to combat the pests and diseases that threaten crop yields every season.
Results of a survey conducted by Feed the Future Innovation Lab for Integrated Pest Management partners at the Tanzanian Agricultural Research Institute (TARI) show that the majority of Tanzanian farmers receive key knowledge on how to manage pests and disease not only from extension personnel, but often from agricultural supply dealers, or agro-dealers. While agro-dealers do carry valuable information, resources and inputs, the survey also shows that many agro-dealers have limited formal knowledge on vegetable production or protective measures for applying chemical pesticides.
To address these gaps, TARI began providing cohesive training to agro-dealers, farmers and extension officers on vegetable production and pest and disease management. Training covers such areas as Good Agricultural Practices (GAPs), Integrated Pest Management (IPM) and safe handling and use of agricultural inputs, including pesticides. Thus far, 500 participants have been trained in the Coast and Morogoro regions. The GAP training in particular helps farmers build capacity in reporting and record-keeping, assessing input quality and crop hygiene, and training in IPM provides information on bio- and botanical pesticides, pruning, developing seedlings in a nursery environment and how to apply pesticides with minimal body exposure.
“Knowing that farmers receive their pest and disease management knowledge from agro-dealers provides us important insight into how to best reach farmers with up-to-date information,” said Dr. Fred Tairo, principal agricultural research officer at TARI-Mikocheni. “If we want farmers to adopt sustainable, climate-smart and productive inputs that might be outside of their typical use, an important pathway to reaching them is through the people that farmers already trust and are familiar with.”
In a group of 69 agro-dealers surveyed, only 49 were registered and licensed to run agricultural shops. The 20 unregistered participants had received no formal training in crop production or pesticide safety and use, and most participants not only had no prior knowledge on how to dispose of expired pesticides, but did not sell bio-pesticides or chemical pesticide alternatives at their shops. Since registering as an agro-dealer can cost nearly $200, TARI is collaborating with the Tropical Pesticides Research Institute (TPRI), a regulatory authority for pesticides in Tanzania, to consider lowering the costs.
TARI and the IPM Innovation Lab are increasing communication through digital platforms to reach more agricultural actors with safe and effective approaches to pest and disease management. A Kiswahili-based (Swahili) WhatsApp group named “Kilima cha Mboga kisasa,” or modern vegetable cultivation, currently shares information with 154 farmers, extension agents and agro-dealers in Tanzania who can use the app to cite crop threats and receive expert management guidance in return.
Participants post a picture or video of the crop problem for immediate diagnosis. Not only do agro-dealers in the group directly learn about farmers’ most pressing problems, but they can use the platform to market agri-inputs, including the IPM products they learn about through the platform.
“Even if members of this group do not necessarily follow up with formal training we offer, this is a low-stakes knowledge-sharing space that they can be a part of and receive guidance from,” Tairo added.
To increase access to information and inputs, the IPM Innovation Lab is also collaborating with Real IPM, a private company based in Kenya that develops low-cost biological and holistic crop solutions available in Kenya and Tanzania. In just one year, the company has provided training to thousands of farmers in seven counties in Kenya by targeting farmer groups, the majority of which are made up of women. Real IPM has developed training manuals on IPM, a WhatsApp group for crop health assistance and a free web portal for diagnosis and IPM recommendations of specific crop threats.
“Our goal is to make IPM solutions more accessible,” said Ruth Murunde, research and development manager at Real IPM. “When you enter a pest or disease into our web portal, those images, diagnosis and IPM recommendations stay posted. We know that many farmers are experiencing similar issues to one another and collective action against crop threats is an effective way to combat them more long-term.”
While technology constraints remain — including smartphone, internet and electricity access — making learning spaces available for a range of crop production actors is critical to adoption of sustainable, effective farming solutions.
Currently, the Real IPM database hosts over 7,000 participants and has collected over 200 infected crop images.
“The Real IPM technical team is actively working to support farmers by providing biopesticides as a solution for mitigating pests and diseases on vegetable crops to ensure sustainable agriculture for smallholder farmers,” added Murunde. “Our information networks help disseminate best practice methods for using those tools.”
Study on tomato leaf diseases classification based on leaf images
Tomato production can be greatly reduced due to various diseases, such as bacterial spot, early blight, and leaf mold. Rapid recognition and timely treatment of diseases can minimize tomato production loss. Nowadays, a large number of researchers (including different institutes, laboratories, and universities) have developed and examined various traditional machine learning (ML) and deep learning (DL) algorithms for plant disease classification.
However, through pass survey analysis, the team found that there are no studies comparing the classification performance of ML and DL for the tomato disease classification problem. The performance and outcomes of different traditional ML and DL (a subset of ML) methods may vary depending on the datasets used and the tasks to be solved. This study generally aimed to identify the most suitable ML/DL models for the PlantVillage tomato dataset and the tomato disease classification problem. For machine learning algorithm implementation, the team used different methods to extract disease features manually. In this study, the team extracted a total of 52 texture features using local binary pattern (LBP) and gray level co-occurrence matrix (GLCM) methods and 105 color features using color moment and color histogram methods. Among all the feature extraction methods, the COLOR+GLCM method obtained the best result.
By comparing the different methods, the team found that the metrics (accuracy, precision, recall, F1 score) of the tested deep learning networks (AlexNet, VGG16, ResNet34, EfficientNet-b0, and MobileNetV2) were all better than those of the measured machine learning algorithms (support vector machine (SVM), k-nearest neighbor (kNN), and random forest (RF)). Furthermore, the team found that, for the dataset and classification task, among the tested ML/DL algorithms, the ResNet34 network obtained the best results, with accuracy of 99.7%, precision of 99.6%, recall of 99.7%, and F1 score of 99.7%.
Lim Chang-won Reporter
Global Plant Protection News 