AI Models

Doktar is a leading provider of AI-powered agricultural solutions. We deliver real-time insights, helping farmers and agribusinesses make data-driven decisions. Our technology can be integrated into IoT devices and applications, providing customized API services for crop health monitoring, pest and disease management, fertilization, and agricultural market intelligence.

• Crop Type Detection
  Uses remote sensing and a state-of-the-art machine-learning algorithm that determines crop types planted in a given area. Combining satellite imagery with ground truth samples gathered through field inspections, the algorithm is trained and optimized to create a model that can detect more than 20 types of crops within a field or an entire country.
• Field Boundary Detection
  Automatically locates and outlines the boundaries of agricultural fields from satellite imagery using deep learning techniques. This process involves feeding our AI model with satellite images and training it to recognize territorial boundaries. The model then applies what it has learned to new imagery and predicts the field boundaries in these images. Automating boundary detection increases efficiency and accuracy compared to the traditional manual methods, allowing institutions to monitor farmers' declarations regarding the boundaries of their fields.
• Pest Detection and Pest Type Recognition
  Uses AI-powered image recognition to detect over 20 types of pests in a given area through a camera on the PestTrap device. It can also estimate the pest population size and notify users promptly. The algorithm is continuously learning and refining its accuracy, even recognizing new pest types within two weeks of appearing.
• Automatic Problematic Zone Detection
  Detects areas with problematic NDVI values in satellite images and displays them on the Health Map while automatically alerting users to potential issues. This AI model removes the complexity of NDVI analysis and the burden of daily field visits. Check out our Plant Health and Growth API service for a complete field scouting solution.
• Early Warnings and Risk Management
  Provides users with a disease risk forecast based on crop type and phenological stage of development. It accurately determines risk levels as high or low by analyzing relative humidity and temperature values.
• Fertilization Program
  Uses crop type and soil analysis to generate a customized and optimized fertilization program quickly and reliably, preventing deficiencies and excessive fertilization while reducing input costs. It recommends the most accurate program for the season and specific parameters.
• Yield Potential Comparison
  Calculates and compares the average NDVI values of two fields planted with the same crop, providing farmers with yield potential indicators. With this information, farmers and agronomists can plan reparative or growth-supporting agricultural activities in areas with low yield potential.
• Live Precipitation and Storm Tracking
  View the live radar map displaying the latest 2-hour movement and path of precipitation and storm clouds, covering over 90 countries worldwide.