Category: Case Studies

Applying AI to Address Phytoplasma Vine Grape Disease

Applying AI to Address a Major Phytoplasma Vine Grape Disease Threat

Applying AI to Address a Major Phytoplasma Vine Grape Disease Threat

Protecting Hungary’s Vine Heritage: AlphaVet and Proofminder Applying AI to Address a Major Phytoplasma Vine Grape Disease Threat. 

AlphaVet, a key player in veterinary and agricultural innovation, invited Proofminder, a leading provider of AI-powered agricultural monitoring, to collaborate on a large-scale vine-monitoring project aimed at detecting symptoms of a major phytoplasma vine grape disease across Hungarian vineyards.  

Hungary’s vineyards, central to the country’s cultural and agricultural heritage, have in recent years faced the spread of Flavescence dorée (FD), one of the most destructive phytoplasma vine grape diseases in Europe. FD poses a serious threat to vineyards: infection may kill young vines outright and greatly reduce the productivity and longevity of older vines. Because there is currently no direct treatment, the only effective protection relies on early detection, targeted removal of infected vines, and strict vector control. Timely identification is critical – once the disease becomes established, it can severely compromise vineyard productivity and threaten long-term vine health.  

How It Works

The process begins with low-altitude drone flights, conducted at approximately 7 meters above the vineyard rows, capturing ultra-high-resolution images of the vines. During the autumn campaign, Proofminder collected 7200 hectares of imagery, total thanks to its extensive network of drone service providers, who responded quickly and supported the nationwide effort. All the data were uploaded to the platform for annotation and AI model development.  

The AI model created by Proofminder is capable of analysing vines at leaf-level resolution to identify potential disease symptoms. Each detected point is automatically geolocated, visualised on a digital map, and compiled into a detailed report for inspector teams. With these health maps, experts can act faster and more effectively, review large areas in less time, and collect targeted samples for laboratory testing. If the lab testing confirms the disease – the plant must be removed from the vineyard to prevent the infection from spreading.

Flavescence dorée (FD) drone monitoring

Capturing ultra-high-resolution images of vineyards

Collaboration and challenges

Proofminder, as the technology partner, was responsible for data collection, image processing, and map generation. Current results indicate that the model can reliably flag vines showing potential symptoms, supporting early intervention and targeted field inspections. At the same time, there is room for optimising flight parameters to further increase efficiency and performance. 

“We need to gather high-quality pictures with reasonable effort. Currently we use 0.2–0.3 GSD resolution, an angled camera, and very low-altitude flights. We will continue experimenting next season to optimise flight time while producing images detailed enough to recognise the disease,” said Levente Simon, Co-founder and CPO of Proofminder. 

AI detecting Flavescence dorée (FD)

Leaf-level detection of Flavescence dorée (FD) by Proofminder AI

Next Steps

With data analysis and model enhancements now underway, the project will soon share its first insights on infection patterns and detection accuracy. The results will help refine both the AI models and operational protocols for future monitoring campaigns. 

Building on this success, AlphaVet and Proofminder are preparing to extend the solution to monitor all 50,000 hectares of vineyards in Hungary, as well as neighbouring wine-producing countries, offering a scalable and proven method for early vine disease detection powered by AI and advanced imaging. 

About Proofminder 

Proofminder is an AI-powered agritech platform that transforms drone and camera imagery into leaf-level insights and prescription maps, enabling growers to reduce input use, cut labour costs, and meet production goals sustainably. Its technology supports leading food and seed production companies worldwide by helping them monitor plant health, detect weeds and plant diseases, plan field operations based on data, and minimise environmental impact. 

About AlphaVet 

AlphaVet is a Hungarian company committed to advancing animal health and agricultural innovation. Through its network, expertise, and partnerships, AlphaVet supports sustainable farming and modernisation of agricultural practices across Hungary. 

 

Interested in learning more or booking a demo? Contact us via the form below.  

Lets talk about your project

Submit the form to get started

Chilean Needle Grass

AI-Powered Detection of Chilean Needle Grass for mine recultivation

AI-Powered Detection of Chilean Needle Grass for Mine Recultivation

Earlier this year, we announced that Proofminder had been selected for the Eureka Innowwide, funded by the European Union, to develop and deploy an AI-powered weed detection system in one of the toughest environments in Australia — mine site rehabilitation.  

This project targets one of the country’s most damaging invasive weeds, Chilean needle grass (CNG) — a declared Weed of National Significance (WoNS) that undermines agriculture, biodiversity, and large-scale land restoration efforts. 

The Challenge: A Persistent and Costly Threat

CNG is no ordinary weed. 

  • It already affects 40 million hectares of Australia and has the potential to spread to 180 million hectares, or 25% of the country’s landmass. 
  • It can reduce productivity by up to 50% and outcompete native vegetation, forming dense monocultures that halt biodiversity recovery. 
  • Each plant can produce 10,000–20,000 seeds per year, many of which remain viable in the soil for over a decade. 

For mine rehabilitation projects, this weed is more than an ecological issue — it’s a compliance and financial risk. 

  • Under the NSW Biosecurity Act 2015, mining leaseholders must prevent and control WoNS weeds. 
  • CNG infestations can delay closure certification, trigger repeated rework and reseeding, and add $2,000–$5,000+ per hectare per year in rehabilitation costs. 

Detection is equally challenging: 

  • Rehabilitation areas can be vast and remote. 
  • Manual inspections are slow, labour-intensive, and prone to missing early-stage plants. 
  • Regulatory sign-off requires consistent proof of control over several years. 

The Solution: High-Resolution Aerial Imaging + AI Detection

Through Eureka Innowwide funding, Proofminder has developed a specialised image capture and AI analysis process tailored for mine site conditions. 

We began by identifying the optimal flight parameters for image gathering — from altitude and speed to camera settings — to ensure our dataset met the precision requirements for training. 

Using a DJI Enterprise drone, we collect imagery at a resolution of just a few millimetres per pixel. This ultra-high detail allows our AI models to detect CNG plants at early growth stages, even when they are visually similar to surrounding vegetation. 

Once the images are captured: 

  • The AI analyses the dataset to locate individual plants with millimetre-level accuracy. 
  • Each detection is logged with exact GPS coordinates. 
  • These coordinates are exportable for any type of treatment — tractor-mounted booms, robotic sprayers, or manual spot application. 

Results to Date: From Field Data to Trained Model 

Since the project began, we have: 

  1. Defined and validated flight parameters that deliver consistent image quality across varied mine site environments. 
  2. Captured and annotated extensive training datasets at a few mm/px resolution. 
  3. Trained and tested the AI model on this data, achieving high detection accuracy across different light and soil conditions. 
  4. Produced precision weed maps fully compatible with existing spraying machinery, robotics, and manual treatment workflows. 
  5. Created regulator-ready reporting outputs that provide verifiable records of weed presence or absence. 

The model is now successfully trained and field-tested, ready for scaling to larger rehabilitation areas. 

Impact Potential

By using this approach, mine operators can: 

✅ Accelerate rehabilitation by removing CNG before it spreads and dominates.

✅ Reduce chemical and labour costsby $2,000–$5,000 per hectare through targeted treatment.

✅ Restore native vegetation and meet biodiversity targets faster.

✅ Prevent off-site spread, protecting surrounding farmland and ecosystems.

✅ Maintain compliance with biosecurity and rehabilitation obligations.

Next Steps 

The next stage of the project will focus on: 

  • Expanding AI model training with seasonal and environmental variability. 
  • Demonstrating consistent performance at scale across multiple mine sites. 
  • Quantifying cost savings and environmental benefits over consecutive monitoring cycles. 

Chilean needle grass is one of Australia’s most persistent and costly invasive weeds — but with the right flight parameters, high-resolution aerial imagery, and AI-powered detection, it can be located early, targeted precisely, and controlled effectively.

Thanks to Eureka Innowwide’s EU funding, this project is demonstrating that even in the most challenging rehabilitation environments, advanced imaging and AI can deliver measurable environmental, economic, and compliance benefits.

👉 Interested in learning more or booking a demo? Contact us via the form below.  

Lets talk about your project

Submit the form to get started

Flowering detection in pineapple

Pineapple Farming: AI and Drones Are Transforming a 33-Month Crop Cycle in Australia

Pineapple Farming: AI and Drones Are Transforming a 33-Month Crop Cycle in Australia

Pineapple is one of the most rewarding — and challenging — crops to grow. With a 33-month production cycle, unpredictable weather, and intensive manual labour, every step matters. From early stand counts to harvest quality checks, precision and timing are everything. 

In Australia, where large plantations and labour shortages collide, Proofminder’s AI-powered monitoring is helping pineapple growers shift from manual estimation to data-driven farming — using drones, ultra-precise AI models, and plant-level insights delivered within hours. 

Why Pineapple Needs a Smarter Approach 

Pineapple cultivation spans nearly three years and involves: 

  • Stand establishment 
  • Weed detection 
  • Flower counts 
  • Fruit classification 
  • Two harvests (around months 20 and 33) 

Traditionally, this meant manual field walks, slow data, and missed opportunities. But today, drone and AI-powered analysis gives growers a full field overview — fast, accurate, and scalable. 

How It Works?

With autonomous drones (like the DJI Dock), high-resolution imagery is collected across thousands of hectares — even in remote areas. That data is processed by Proofminder’s platform and returned as: 

  • Heatmaps
  • Plant-level counts
  • AI-based insights — all in under a few hours 

Key Use Cases in Pineapple Production

1. Fruit Counting

Growers can now accurately predict yields early in the cycle — even when fruits are hidden under foliage. 

  • Plan pricing, supply, and budgets with confidence
  •  Time harvests for peak market value
  •  Detect fruit issues early to reduce losses

2. Flower Counting

Pineapple flowering is rarely uniform — Proofminder helps track it accurately. 

  • Synchronise harvests for better quality and fewer field losses
  • Improve labour efficiency with better scheduling
  • Support yield forecasting with precise counts
  • Supply chain smoothing 
Flower detection and counting in pineapple

Image 1. Flower detection and counting in pineapple by Proofminder AI

3. Stand Counting

Edge scouting misses most of the field — AI doesn’t. 

  • Identify gaps and poor establishment across the whole crop 
  • Optimise use of water, fertiliser, and plant protection
  • Plan interventions with clarity 
Pineapple plant and fruit counting with AI

Image 2. Proofminder AI counting pineapple plants

What Growers Gain 

✅ Up to +10% yield improvement 
✅ Hundreds of hours saved on manual scouting 
✅ Up to 80% less chemical use 
✅ Higher crop quality, less waste 
✅ Stronger resilience to environmental and market challenges 

What’s Next?

Weather damage analysis, weed detection, and other advanced tools are already in development to further improve pineapple productivity in Australia and beyond. 

 

👉 Interested in learning more or booking a demo? Contact us via the form below.  

Lets talk about your project

Submit the form to get started

Cotton field - density analysis with Proofminder

Smarter Cotton Farming at Scale — Powered by Plant-Level AI

Smarter Cotton Farming at Scale — Powered by Plant-Level AI

The Australian cotton industry is a significant player in the global market, known for its high-quality production and substantial export contributions. Australia is among the world’s top four cotton exporters, with over 90% of production shipped internationally — primarily to China, Vietnam, and Bangladesh. 

However, the industry is navigating a set of growing challenges: 

  • Water Security: Reliance on irrigation makes cotton vulnerable to drought, increasingly worsened by climate volatility. 
  • High Costs: Biotechnology, energy, and machinery contribute to rising input costs. 
  • Market Pressure: Brazil’s expanding cotton exports are driving down global prices. 
  • Synthetic Fibre Competition: Cheaper options like polyester are cutting into global cotton demand. 
  • Trade Risks: Tensions with key buyers, especially China, pose long-term uncertainties. 

To stay competitive, the industry is investing in sustainability and technology — positioning Australian cotton as a premium, environmentally responsible product. A key part of this transformation is precision agriculture. 

Smart Cotton Monitoring

Australian cotton farms often span thousands of hectares. Monitoring these vast areas manually is not only time-consuming — it’s nearly impossible to do consistently. This limits the grower’s ability to detect early issues, make timely decisions, and optimise productivity across the entire season. 

Proofminder’s AI-powered crop monitoring solves this by combining high-resolution drone imagery with advanced AI models — delivering plant-level insights in under a few hours, across entire fields. 

Here’s how Proofminder supports growers through the entire cotton cycle: 

1. Emergence Analysis to Support Replanting decisions  

 In the early stages, uniform emergence is critical. Patchy germination or early plant loss due to waterlogging, compaction, or poor seedbed prep can reduce final yields. 

Proofminder delivers: 

  • Plant density maps to assess stand establishment 
  • Gap detection across the full field — not just field edges 
  • Replanting support through zone-specific insights 

Outcome: faster, data-backed decisions before the crop gets beyond recovery. 

Precise cotton stand counting by AI

Image 1. Precise cotton stand counting by Proofminder AI

Cotton field - density analysis with Proofminder

Image 2. Cotton plant density analysis, or the number of plants per hectare

2. Weed Detection and In-Season Monitoring

Weed pressure continues to be a top input cost in cotton. With Proofminder: 

  • Weed patches are automatically detected and mapped 
  • Prescription maps are generated for spot spraying or automated removal
  • Stress indicators help flag areas with pest damage or nutrient issues 

Outcome: growers can take swift, targeted action while reducing herbicide use and protecting crop health. 

3. Cotton Classification by Size

As the crop matures, understanding how plants are developing is key to yield forecasting and harvest planning. 

Proofminder uses plant size classification to: 

  • Track vegetative growth and canopy uniformity 
  • Highlight underperforming zones 
  • Support block-level yield predictions and mid-season input decisions 

Outcome: a more accurate, data-backed understanding of crop potential across every hectare. 

Cotton plant analysis - classification by size

Image 3. Cotton field analysis – small plant area

Cotton field analysis – bigger plant area

Image 4. Cotton field analysis – bigger plant area

Automated Insights That Integrate Seamlessly

All reports from Proofminder are built for real-world farm operations. Our maps and insights are: 

  • Compatible with traditional sprayers, robotics, and autonomous machinery 
  • Exportable to existing farm management software and GPS systems 
  • Easily layered with on-farm sensor and IoT data for deeper context 

This means no extra hardware, no disruptive workflows — just faster, smarter decisions supported by tech that fits your farm. 

✅ Real world impact

  • Up to +10% potential yield improvement 
  • Up to 80% less chemical use via precision input application 
  • Hundreds hours saved on manual scouting per field 
  • Reduced input waste, better resource allocation 
  • Stronger forecasting and risk management throughout the season 

Want to See It in Action? Simply contact us via the form below.

 

Lets talk about your project

Submit the form to get started

tomatoe yield

Boosting Tomato Yields with AI-Powered Transplant Monitoring

Boosting Tomato Yields with AI-Powered Transplant Monitoring

Transplanting tomatoes is a critical and delicate phase in the growing cycle. Young seedlings are highly sensitive, and even minor damage during transplanting can significantly reduce final yields. But what if you could detect and respond to issues within 24 hours—before they affect your bottom line?

With Proofminder’s AI-powered transplant monitoring, tomato growers can now take the guesswork out of early-stage crop management.

How It Works

Using drones or vehicle-mounted cameras, Proofminder captures ultra-high-resolution imagery of your tomato fields immediately after transplanting. These images are processed by our advanced AI models to deliver timely, field-wide report on plant health.

tomatoe field monitoring

Image 1. Drone capturing post-transplant images for analysis 

Our system automatically identifies and quantifies:

  • Healthy transplants
  • Heat-damaged plants
  • Dead or failed seedlings
AI-generated transplant health map from Proofminder showing tomato field stress detection

Image 2. Proofminder heat map visualising healthy and damaged tomato transplants for early crop performance analysis.

Why It Matters

Instead of discovering problems too late, growers can:

  • Replant underperforming zones before it’s too late
  • Reduce yield loss by catching stress early
  • Prioritise resources and fieldwork based on accurate maps

This technology was successfully used by one of the world’s largest tomato producers, proving its value at scale. Not only does it save labour costs associated with manual scouting, but it also increases yield security and boosts ROI by enabling faster, smarter field interventions.

Turning Data into Results

By turning raw field data into actionable insights, Proofminder helps tomato growers:

  • Save time
  • Reduce input waste
  • Improve profitability
  • Strengthen crop resilience

With AI-powered transplant monitoring, tomato farming becomes smarter, faster, and more sustainable.

🍅 Ready to take your tomato yields to the next level? Let’s talk about how Proofminder can support your season!

Lets talk about your project

Submit the form to get started

dropped fruit counting in the orange orchard

Citrus Greening: A Threat to Citrus Production and How to Combat It

Citrus Greening: A Threat to Citrus Production and How to Combat It

Citrus greening, also known as Huanglongbing (HLB), is a bacterial infection that poses a severe threat to citrus species worldwide. It is often a death sentence for infected trees, leading to significant yield losses and economic damage. First identified in China in the early 1900s, the disease has rapidly spread to major citrus-producing countries, including Brazil and the United States, where it has had a devastating impact on the industry, affecting growers, producers, and consumers alike. Although various strategies exist to control the spread of the disease, the question remains: How can innovative technologies like Proofminder help?

Overview of Citrus Greening Disease

Citrus greening was first documented in 1919 in Guangdong Province, South China. The disease is caused by three species of the bacterium Candidatus Liberibacter: Ca. L. asiaticus, Ca. L. africanus, and Ca. L. americanus, with the Asian form being the most widespread and destructive. 

The pathogen is transmitted by the Asian citrus psyllid (Diaphorina citri), a small flying insect. The bacteria reside in the phloem, the nutrient-conducting tissue of the tree, and are spread when the psyllid feeds on the infected phloem. Once the psyllid becomes a carrier of Ca. L. asiaticus, it can infect other trees, leading to a rapid and widespread transmission of the disease. 

Symptoms and Effects on Citrus Trees

Citrus greening primarily attacks the phloem of the tree, disrupting nutrient transport. Symptoms include premature fruit drop, misshapen and bitter fruit, stunted growth, and sparse canopies. Over time, infected trees become unproductive and may die within a few years, severely impacting citrus yields. 

Strategies for Controlling Citrus Greening

Given the severity of citrus greening, various strategies have been developed to control its spread, although finding long-term solutions has been challenging due to the elusive nature of the bacterium. Scientists have struggled to grow the bacteria in laboratory conditions, making research and development more difficult. 

Protecting Against the Bacteria and the Psyllids

Orange farmers utilize a mixture of strategies to minimize the damage caused by citrus greening. 

  • Chemical Control: Chemical pesticides can help control psyllid populations, but insects often develop resistance, and there are environmental concerns associated with their use. Alternatively, biological control methods, such as using the wasp Tamarixia radiata, have shown promise. The wasp lays its eggs inside psyllid larvae, killing them as the wasp larvae develop. Additionally, RNA interference techniques are being explored, involving the use of viruses to introduce HLB-killing molecules into the trees. 
  • Bacterial Protection: Farmers have attempted to combat the issue with antibiotics, but the pathogen often resides deep in the roots, beyond the reach of these treatments, raising concerns about human health risks. Plant breeding has also seen some success, such as with the HLB-tolerant variety “Super Belle,” which was developed by chance. However, creating an HLB-resistant variety remains a priority. Due to the slow pace of traditional plant breeding, researchers are turning to gene editing and other advanced biotechnological techniques.  

Prevention and Early Detection 

Prevention is a critical part of battling citrus greening. In China and Brazil, successful prevention strategies involve detecting and removing infected trees and monitoring and controlling psyllid populations. 

How Proofminder Can help producers Mitigate the effects of Citrus Greening 

Use Case 1: Dropped Fruit Counting 

Premature fruit drop is one of the most critical symptoms of citrus greening, leading to significant income loss for growers and indicating the presence of infected trees. To address this issue, Proofminder has developed an AI-driven model that leverages visual data collected from the fields. This technology enables precise counting of dropped fruit across vast orange plantations encompassing thousands of hectares of land. By analyzing high-resolution images, the AI model detects and quantifies premature fruit drop with remarkable accuracy. The dropped fruits are marked with GPS coordinates, which are provided to workers who can collect them and send them to be juiced. If picked up – thus juiced – in time, dropped oranges still can have economic value.  

dropped fruit counting in the orange orchard

Figure 1 – Proofminder’s AI model for dropped fruit counting in the orange orchard

Use Case 2: Tree Counting and Classification 

Proofminder offers advanced tools for tree counting and classification, categorizing trees in the orchard into living trees, dead trees, and seedlings, while also calculating the number of missing trees. This provides growers with a clear overview of their orchard’s status, helping them assess the extent and location of infestations and reduce spraying costs. Additional benefits include: 

  • Quantified Vegetation Assessment: Based on spectral indexes (stress, water, nitrogen), this assessment helps detect irrigation blockages, provides inputs for regional management, and isolates problematic areas. 
  • Tree Volume Measurement: Canopy size and height measurements are used to calculate the necessary amounts of chemicals. 

Proofminder’s technology delivers precision, with a target accuracy of 95%. Data is accessible via an online platform, offering multiple access points, zoomable images, and dynamic analysis layers, along with export options in Excel, CSV, and geospatial formats (KML, Shapefile).

orange tree classification by size

Figure 2 – Proofminder’s AI model for orange tree counting and classification

Conclusion

Citrus greening is a serious disease that has led to a 20% decrease in the fresh citrus fruit market and a 70% decline in orange production for juice and other products in the United States. If you are interested in learning how Proofminder can assist in detecting infected trees and preventing the spread of HLB in your orchard, don’t hesitate to reach out to us. 

Lets talk about your project

Submit the form to get started

Cabbage plant counting by Proofminder AI model

AI-Powered Precision: Revolutionizing Cabbage Counting

AI-Powered Precision: Revolutionizing Cabbage Counting

Accurately counting cabbages is crucial for precision agriculture, as it enables growers to estimate yields, manage and allocate resources, plan labor and logistics, and control quality effectively. Traditional, manual counting methods are labor-intensive, time-consuming, and prone to errors, especially in large fields. 

The challenge

A large-scale cabbage grower in Southern Africa, managing fields spanning several hectares, faced significant challenges with manual cabbage counting. The process was both labor-intensive and error-prone, leading to inefficiencies in planning and resource allocation. To address these issues, the grower sought an innovative solution to streamline the counting process.  

Renier Jonker and Angelo De Andrade, owners of Precision AG in Southern Africa and representatives of Proofminder in the region, recognized the potential for drone technology to revolutionize cabbage counting. They proposed using drones and AI to streamline the process, addressing the challenges of high labor costs, limited infrastructure, and the growing demand for precision farming in Africa.  

This use case is a prime example of how technology can be tailored to meet the specific needs of growers in Africa, where traditional methods often fall short. Precision AG’s insight into the local market and its ability to adapt global technologies to local conditions played a vital role in the success of this innovation.  

The process

A large-scale cabbage grower in Southern Africa, managing fields spanning several hectares, faced significant challenges with manual cabbage counting. The process was both labor-intensive and error-prone, leading to inefficiencies in planning and resource allocation. To address these issues, the grower sought an innovative solution to streamline the counting process.  

Renier Jonker and Angelo De Andrade, owners of Precision AG in Southern Africa and representatives of Proofminder in the region, recognized the potential for drone technology to revolutionize cabbage counting. They proposed using drones and AI to streamline the process, addressing the challenges of high labor costs, limited infrastructure, and the growing demand for precision farming in Africa.  

This use case is a prime example of how technology can be tailored to meet the specific needs of growers in Africa, where traditional methods often fall short. Precision AG’s insight into the local market and its ability to adapt global technologies to local conditions played a vital role in the success of this innovation.  

1. Image Gathering:  

Using a DJI Mavic 3 M drone equipped with a high-resolution camera, detailed images of the fields were captured at an impressive 0.5 GSD (Ground Sampling Distance). This precision was essential for accurately identifying and counting cabbages, as well as distinguishing their difference in size. This high-resolution imagery ensured that even small variations in size and spacing were detected.  

2. AI Model Development: 

Proofminder developed an AI model tailored to the needs of cabbage growers. The model was trained to count the vegetables, categorize them by size (small, medium, large), and assess potential yield with remarkable accuracy, taking into account factors such as plant health, growth patterns, and environmental conditions. 

3. Report Generation: 

Within days, the AI model generated a hyper-accurate report, providing the grower with detailed insights into crop status, including cabbage counts, size distribution, and yield estimates. This information, accessible via a web application, enabled informed decisions about harvesting, marketing, and resource allocation. 

Cabbage plant counting

Figure 1 – Proofminder platform: Cabbage plant counting 

Cabbage plant counting by Proofminder AI model

Figure 2 – Cabbage plant counting by Proofminder’s AI model

Challenges and solutions

1. Environmental Conditions: 

  • Challenge: Changes in weather, such as strong winds and cloud cover, can affect the quality of drone-captured images. 
  • Solution: Drone flights were scheduled during optimal weather conditions, and sensors were calibrated to improve image consistency. 

2. Field Size and Terrain: 

  • Challenge: Large and uneven terrains posed challenges for consistent image capture, especially in areas with dense vegetation and irregular plot shapes.
  • Solution: Planning drone flights using advanced flight planning software ensured complete field coverage, with overlapping images to maintain accuracy. 

3. Data Processing: 

  • Challenge: High-resolution images and large field sizes resulted in vast amounts of data, which require efficient processing. 
  • Solution: Proofminder’s AI model and processing infrastructure was optimized for large datasets, ensuring timely delivery of accurate reports. 

These challenges underline the importance of planning and adaptability when integrating drone technology into agriculture. The success of the project is a testament to the team’s ability to anticipate and overcome these, ensuring that the final solution met the grower’s needs effectively.  

Outcome, Results, Insights

The analysis of drone images provided precise yield estimates, aiding in better planning, resource management, and market alignment. Additional insights included details on field health, plant health, and weed detection, leading to improved crop management and efficiency. 

The project’s economic impact is significant: 

  • Increased Profitability: Higher revenue through accurate yield estimates. 
  • Cost Savings: Reduced labor and resource costs. 
  • Improved Efficiency: Streamlined operations and better resource use. 
  • Market Competitiveness: Enhanced ability to meet market demands and adjust pricing. 

Feedback from the grower highlighted the system’s transformative impact in the following areas:  

  • An increase in accuracy in yield estimates and size classification, which helped streamline  planning and resource management.  
  • Improvement in efficiency regarding field management and harvest planning, thanks to reduced labor costs and time saving.  
  • Enhanced insights into plant health and weed detection allowed for timely interventions and better overall crop management.  

Who Can Benefit?

This AI-driven technology can be applied to other vegetable crops, row crops like corn or soybeans, and specialty crops, offering detailed insights and improved efficiency across various agricultural sectors. 

Ready to revolutionize your crop management? Leverage AI-powered solutions and drone technology to optimize yield, reduce costs, and make smarter decisions for your farm. 

Contact Proofminder today to discover how our innovative tools can transform your agricultural operations! 

Lets talk about your project

Submit the form to get started

Onion umbels counting by AI

AI-Driven Innovations for Onion Seed Producers

AI-Driven Innovations for Onion Seed Producers

Onion seed production is a sophisticated process that faces numerous challenges, including climate variability, pests, diseases, and the necessity for sustainable agricultural practices. A critical aspect of this process is accurately counting onion umbels, as this allows growers to estimate the potential yield of their crops and devise more effective crop management strategies. 

Onion umbels counting - use case background

An onion producer managing a 7-hectare field in Zaragoza, Aragon, sought an innovative solution to enhance the accuracy of yield evaluations. The producer was facing difficulties in accurately identifying umbels and distinguishing them from other objects in the field, such as rocks. 

Luis Valdez, our representative in Spain, led the project, working closely with the producer to understand their specific challenges and needs. His expertise in agricultural technology and deep understanding of local farming practices were instrumental in tailoring the solution. 

The process

To address these challenges, images of the field were collected using a DJI Phantom 4 RTK drone, providing high-resolution data at 0.5 GSD. Proofminder then utilized this data to train an AI model capable of: 

  • Counting onion umbels. 
  • Classifying umbels by size (small, medium, and large). 
  • Delivering a hyper-accurate report to the producer via a web application.
Onion umbels counting by AI

Figure 1 – Proofminder’s AI model for onion seed growers

Onion umbels counting by size

Figure 2 – Onion umbels counting and classified by size (small, medium, and large)

Challenges in the innovation process

During the development phase, it became apparent that the AI model needed to differentiate between onion umbels and white rocks in the field. To solve this, the model was fine-tuned to recognize rocks separately, achieving an impressive accuracy rate of over 95%. This additional feature not only enhanced the accuracy of umbel counts but also contributed to safer cultivation practices by helping the grower manage the presence of rocks in their field. 

The results

  • Accurate Yield Assessment: The technology enables a more precise evaluation of crop yields. 
  • Cross-referencing Information: The model’s accuracy can be validated through sample counts or during harvest. 
  • Evaluation of Productive Potential: The field can be zoned more effectively, allowing for a more accurate assessment of its productive potential. 
  • Additional Insights: The model provides additional data about the field and plants, aiding in better decision-making, such as: 
  • Onion umbels were categorized by size (small, medium, large).  
  • Results allow the grower to compare the yield with that of previous crop years.  
  • If the field is rented, it allows the grower to make an educated decision on whether to discard or ratify the work with the same field. 
  • Digitized Information: All information is presented in a digitized format, easily accessible via a web application. 

A Look Ahead

This AI model is versatile and can also be applied to other crops such as garlic, leek, watermelon, melon, pumpkin, carrot, parsley, fennel, coriander, and dill. 

As the agricultural landscape continues to evolve, the integration of AI technologies is no longer a luxury but a necessity for competitive and sustainable farming. By adopting innovative solutions like AI-driven umbel counting models, onion seed producers can optimize their yields, reduce risks, and make data-driven decisions that enhance overall productivity. 

If you’re a grower looking to elevate your farming practices with cutting-edge technology, don’t hesitate to explore how Proofminder’s AI models can be tailored to your specific needs. 

Contact us today to learn more about how we can help you achieve greater accuracy and efficiency in your agricultural operations.  

Lets talk about your project

Submit the form to get started

wildlife damage assessment

Wildlife Damage Analysis and Vole Hole Detection in the Field

Wildlife Damage Analysis and Vole Hole Detection in the Field

Voles have always posed significant challenges for growers due to their feeding habits on the roots, stems, and seeds of crops, which can severely damage plants and reduce yields. When vole populations become too large, they can decimate entire fields by consuming large amounts of vegetation and undermining plant stability. Their burrowing activities disrupt the soil structure, making it difficult for crops to establish strong root systems. Additionally, voles can spread diseases and parasites to plants, further compromising crop health. An overabundance of voles can lead to severe economic losses for farmers, as the destruction of crops directly translates to decreased productivity and profitability. 

This problem is growing worse each year. Aerial imagery shows the increasing extent of this issue, making the matter more and more urgent, especially in Slovakia, where most farmers are seeking a solution for these critters. Precisely pinpointing where the burrows are can be a powerful tool. During Techagro, many farmers expressed interest in burrow detection. 

When Proofminder was presented to Slovakian and Czech agricultural organizations, one of the most frequently asked questions was whether we could help identify voles and their burrows. Our first real use-case came from a leading player in Slovakia, who requested a test of Proofminder’s capabilities on 20 hectares of wheat near the city of Piešťany. The process is described below. 

First, we had to choose good weather conditions for the mission—not too windy (which would affect drone stability) and sunny. The best pictures can be taken when there is a thin layer of clouds, but sunny conditions without scattered clouds will also do (scattered clouds can create unwanted shadows in photos, which may result in false positives, especially during burrow detection). We had sunny conditions without clouds and very little wind; image gathering started at 10 AM. The drone used was a Mavic 3 Enterprise, with six spare batteries and a charging station. The first part (10 hectares) took around one hour to fly, with a 0.5 cm/px resolution at a flying altitude of 16 meters. The second half was flown at an altitude of 13 meters, with a 0.42 cm/px resolution, taking slightly over an hour. 

Once the mission was completed, the captured images were uploaded to Proofminder’s environment, and the work began. After stitching the photos into an orthomosaic, we corrected issues with shadows to avoid false burrow detection and let Proofminder’s advanced algorithm work its magic. The results are shown below.  

We confirmed a 97% success rate. We performed a visual check on-site comparing Proofminder results with our observations. The detection of the burrows was meticulous. After the harvest, we will continue with burrow detection to fine-tune the process. 

These results can help farmers create a comprehensive plan to combat the vole problem. There are multiple methods to reduce vole populations, and Proofminder provides all the data needed to achieve this goal. 

wildlife damage assessment

Figure 1 – Wildlife damage to the wheat field

vole hole detection by Proofminder AI model

Figure 2 – Vole holes detected by Proofminder’s AI model

Proofminder can help address this significant problem by: 

  1. Accurately detecting the exact location of vole holes in the field. 
  2. Estimating damages. 
  3. Helping plan the most effective plant protection strategy against voles. 
Wildlife damage AI analysis and report

Figure 3 – Vole holes detected by Proofminder’s AI model 

A Few Strategies for Managing Vole Populations: 

  1. Habitat Modification: Remove weeds, tall grasses, and debris from field margins and around crops to eliminate vole shelters and reduce their food sources. 
  2. Crop Rotation: Regularly changing the types of crops grown can disrupt the voles’ feeding patterns and reduce their populations. 
  3. Barriers: Install physical barriers like wire mesh around fields to keep voles out. 
  4. Biological Control: Encourage natural predators such as owls, hawks, snakes, and foxes by creating suitable habitats for them. Installing perches or nest boxes can attract birds of prey. 

These integrated pest management strategies help to keep vole populations under control, protecting crops and maintaining agricultural productivity. 

Contact us to receive an accurate analysis of your field, assess yield loss and improve yield. 

Lets talk about your project

Submit the form to get started

Rock detection on the field

AI-Powered Rock Detection for Safer Cultivation

AI-Powered Rock Detection for Safer Cultivation

Rocks in the arable land can pose significant challenges during cultivation, and as such, their removal from is advantageous for several reasons: 

  1. Rocks can obstruct agricultural machinery, leading to damages and thus, increased maintenance costs. 
  2. By clearing the land of rocks, farmers can achieve a smoother soil surface, making plowing, planting, and harvesting operations more efficient and less labor-intensive. This can also result in a more uniform seedbed, promoting better crop germination and growth. 
  3. The removal of rocks minimizes the risk of crop damage during mechanized farming, thereby potentially increasing yield and productivity. 
  4. A cleared field is easier to irrigate and manage, leading to better overall farm management and efficiency.
Rocks detected by Proofminder AI model

Figure 1 – Rocks detected on the field by Proofminder AI model 

Rock detection on the field

Figure 2 – Rocks detected by AI on the agricultural field

As can be seen on Figure 2, a significant amount of rocks were detected on the field. It is absolutely recommended to remove the stones, to increase the safety of cultivation. 

Rock spotting technology

Figure 3 – Rocks detected by Proofminder AI model

Rock identification on the arable land

Figure 4 – High-accuracy rock detection

Proofminder offers a solution in the form of high-accuracy rock detection. With the help of our AI model, the stones can be identified, and their location pinpointed with precise GPS coordinates. Thanks to these insights, a route can be generated, and the rocks can easily be collected so that damage to machinery is minimized, if not eliminated.  

Lets talk about your project

Submit the form to get started