Greenhouse Monitoring System Based on Image Spectral Data

Table of Contents

  1. Introduction
  2. Problem Statement
  3. Our Solution
  4. System Modelling
  5. Solution Architecture
  6. Software Design
  7. Machine Learning Model
  8. Team
  9. Links

Introduction

Greenhouse Monitoring System provides a platform to manage the Greenhouse by tracking the phases of plant harvest, identifying any plant disorder and tracking the plant growth by using the image spectral data of plants. The system is basically considered the key problems plant diseases, huge harvest wastage and unnecessary expensive maintenance in a greenhouse. So this system will make a high positive impact on maximizing the harvest and reduce maintenance cost in Greenhouses.

Introduction

Problem Statement

Although the environmental conditions of plants are controlled, the temporal effects like temperature, humidity may not evenly balanced for each crop. Therefore, plants respond differently under those unbalanced environmental conditions. And, plants can have different kind of disorders. It leads to production failures in greenhouses. As well, the crop yield may not be harvested at the right moment. Because of that there would be a huge harvest wastage.

In current greenhouses, the workers continuously observe the plant growth. In that case, workers will be tired and labor system would be inefficient. So, there would be an unnecessary higher cost for maintenance.

Problems

Our Solution

Plant diseases, huge harvest wastage and unnecessary expensive maintenance have been observed as the major issues in current Greenhouse Systems. These key problems simulated the development of image analysis and computer vision methods. That’s how the “greenhouse monitoring system based on image spectral data” came to the stage.

Our_solution

System Modelling

Functional Requirements

Non-Functional Requirements

Use Case Diagram

usecase_diagram

Solution Architecture

System Functionalities

  1. Extract images from the video file.
  2. Extract features from images.
  3. Store those images and features in a database.
  4. Data processing and data analysis using ML Model
  5. Data Visualization
  6. Data Prediction and generate reports

High_level_diagram

System Overview with Technology Stack

Technology_stack

Software Design

User’s Application Data Flow

app_data_flow

User Interfaces

  1. Login
    • Greenhouse workers can login to the system by entering username and password. login
  2. Dashboard
    • User can monitor the overall status of greenhouse
    • User can see what plants have problems in the greenhouse layout panel. and if user clicks the layout it will navigate to the Overview User Interface. dashboard
  3. Overview
    • User can monitor all plants at once. overview
  4. One Plant Overview
    • User can monitor only one plant
    • If user needs to see the diseases, growth and harvest status of that plant, he can click those options in the user interface one_plant_overview
  5. Leaf Diseases
    • If the plant has any disease it will be shown in this interface. diseases
  6. Plant Growth
    • The plant growth can be monitored here. growth
  7. Crop Harvest
    • The plant harvest status can be seen here. harvest
  8. Predictions
    • The predictions of diseases, growth, harvest of each plant can be monitored here. predictions
  9. Reports
    • The reports of the greenhouse system status can be seen and downloaded from here. reports

Machine Learning Model

Leaf Disease Detection

Leaf Disease Detection Model Architecture

leaf_diseases_model

Crop Harvest Stage Prediction

crop_harvest_ml_model

Test Results

Leaf Disease Detection

test_results_1

Team

Developers

Scrum Master

Product Owners