Crop Yield AI — Smart Agriculture Prediction System

AI Model Active

Predict Crop Yield
with AI Intelligence

Upload satellite or drone imagery to get instant yield predictions and smart crop recommendations.

AI-Powered

Instant Results

Location-Aware

STEP 1

Upload Imagery

Select satellite or drone photos. Supports PNG, JPG, TIFF, and WebP.

STEP 2

AI Analysis

Our model extracts vegetation indices and environmental data.

STEP 3

Get Results

Receive yield predictions and top crop recommendations.

No Image? Try an Example

Try This

Example 1

Try This

Example 2

Try This

Example 3

Upload & Analyze

Click to upload or drag & drop

PNG, JPG, JPEG, TIF, TIFF, WEBP

Maximum 15 images per upload

No files chosen

Prediction History

IMAGES ANALYZED

0.0

AVG YIELD (T/HA)

No predictions yet. Upload an image to get started!

User Profile

Full Name

Email Address

Organisation / Institution

Select your organisation...

Other...

Role

Preferences

Yield Unit

Choose your preferred measurement unit for yield values

Notifications

Analysis Complete Alert

Show a notification when image analysis finishes

Low Yield Warning

Alert when predicted yield is below threshold

Data & Storage

Prediction History

Loading storage info...

Reset All Settings

Reset all preferences and profile data to defaults

Quick Start Guide

Upload Your Image

Click Upload & Analyze in the sidebar. Drag and drop — or click the upload area — to select satellite or drone imagery. Supported formats: PNG, JPG, JPEG, TIF, TIFF, WebP. You can upload multiple images at once.

Run AI Analysis

Click PROCESS ANALYSIS. The AI model extracts vegetation indices (NDVI, EVI) and colour features from the image to estimate crop health and yield potential.

Add GPS Coordinates (Optional)

For richer predictions, click Add Location on any result card and enter your field's latitude and longitude. This fetches live weather data and soil type for your area.

Review Results & Recommendations

Each result card shows the Predicted Yield (T/Ha), a Yield Class (Good / Fair / Poor), top recommended crops, and an environment snapshot with weather and soil data.

View & Export History

All predictions are saved in History. Click any history item to expand the full result. Use Export to download a CSV of all predictions.

Frequently Asked Questions

Multispectral satellite imagery or high-resolution drone photos (above-canopy view) give the most accurate results. The model works best with images taken during the growing season showing clear crop rows or vegetation cover.

The AI analyses vegetation-related colour features and indices from the uploaded image. These values feed into a trained regression model to produce a yield estimate in Tonnes per Hectare (T/Ha). If GPS coordinates are provided, weather and soil data are also factored in.

Good — yield is above the expected average for the region and crop type. Fair — yield is within an average range. Poor — yield is below average, suggesting stress factors such as drought, disease, or poor soil.

Crop recommendations are generated by a separate Random Forest classifier trained on Malaysian agricultural data. It considers the image-derived environmental features (and GPS-based weather/soil data when available) to suggest the top 3 most suitable crops for your plot.

Yes! Select multiple files in the upload dialog or drag multiple images at once. Each image is analysed independently and results are ranked from highest to lowest predicted yield on the Results page.

Uploaded images are temporarily stored on the server in an images/ folder for the duration of your session to allow location-based re-analysis. Prediction history is stored locally in your browser's localStorage and is not sent to any external server.

Enter the latitude and longitude of your field in decimal degrees. For example, Kuala Lumpur is approximately 3.1390, 101.6869. You can find coordinates on Google Maps by right-clicking on your field location.

About This System

AI Model

Random Forest Regressor + Classifier trained on Malaysian multi-crop agricultural datasets with image feature extraction pipeline.

Weather Data

7-day forecast sourced from the Open-Meteo API using your GPS coordinates for temperature, precipitation and humidity.

Soil Data

Soil texture and pH data retrieved from the SoilGrids / ISRIC World Soil Database using coordinate-based lookup.

Version

Crop Yield AI v2.0 — Final Year Project (FYP). Built with Python Flask, scikit-learn, and Vanilla JS.

Predict Crop Yield with AI Intelligence

Upload Imagery

AI Analysis

Get Results

No Image? Try an Example

Example 1

Example 2

Example 3

Upload & Analyze

Click to upload or drag & drop

Prediction History

User Profile

Preferences

Notifications

Data & Storage

Quick Start Guide

Frequently Asked Questions

About This System

Add Location Data

Predict Crop Yield
with AI Intelligence