Efficient Utilization of Fertilizer in Sustainable Cotton Production Using IoT

United Nations has estimated that the world population will increase to 8.3 billion in 2030 and 9.3 billion in 2050 due to which the demand for food, fibre and fuel continues to grow, increase in such demands need to be met by either increasing the areas of land that may not be suitable for crop pro

Project Title

Efficient Utilization of Fertilizer in Sustainable Cotton Production Using IoT

Project Area of Specialization

Internet of Things

Project Summary

United Nations has estimated that the world population will increase to 8.3 billion in 2030 and 9.3 billion in 2050 due to which the demand for food, fibre and fuel continues to grow, increase in such demands need to be met by either increasing the areas of land that may not be suitable for crop production or else by increasing yield. Cotton sector seeks to maximize yields and needs to minimize costs. ICT in agriculture is the solution to the growing need of cotton crop. Nitrogen (N) is a major element for cotton plant growth and is a radical part of chlorophyll (Ch). Chlorophyll and Nitrogen are dependent on each other. The deficiency of nitrogen can affect photosynthetic processes, as well as water uptake, yield, leaf size, fiber quality, and number of nodes. Furthermore, nitrogen deficiency appears as stunted plants with a yellowish-green leaf color and reduces size. To acquire effective cotton crop yield prediction, it is important to make precise fertilizer decisions. Nitrogen status in plants is usually identified from leaf color by naked eye of crop expert or through leaf destruction lab tests. Many methods are developed to find N status of plants using image processing. Although existing methodology played considerable role to predict the existing nitrogen amount its deficiency or accessibility as well as proposed the advisory for sustainable growth, yet it requires precise consideration to improve the yield. Proposed methodology will be transferred the destructive manual process into a non-destructive using color detection sensors (Arduino TCS230/TCS-3200) and image processing techniques to get more accurate results, Data of cotton leaves of different varieties will be acquired by applying RGB models the color and texture features of cotton leave will be extracted. The main advantage of proposed methodology is to make the use of fertilizer efficient which save resources as well as environment from fertilizer side effects (eco-friendly).

Project Objectives

• To Identify the amount of nitrogen and predict plants stress level due to nitorgen deficiency in cotton plant
• To predict how much more nitrogen is required for its healthy growth, that leads to efficient fertilizer use

Project Implementation Method

Our procedures consist of three phases that are outlined below.

1. Dataset collection
2. Grading
3. Preprocessing of images

Dataset collection

TCS-3200/TCS230 sensors will be used to get RGB value of leaves from real-time cotton leaves from cotton fields, The dataset will also be taken from real-time images of cotton leaves from cotton fields. Leaf condition in an image is a single leaf image collected by using digital camera from sowing to picking stage of cotton leaves.

Grading

   We grade this data set according to the amount of nitrogen present in the leaf. So, we have N-Completeness, N Deficiency, and N Accessibility grades based on the cotton leaf color.

   Pre-processing of images.

   The pre-processing of images involved different section.

• Original images
• Rescaling of images
• Normalization
• Texture feature
• Feature Extraction and Optimization

After preprocessing of images, the features (e.g., texture features) will be extracted from images these features will be optimized to prepare an optimized dataset of each class. To improve the performance of the model, it can sometimes be desirable to reduce the number of input variables. This can reduce the cost of modeling and, in some cases, reduce the computational burden on the model

  Languages

    Python, C++

  Framework

    Google Collab, Dev C++

Benefits of the Project

Proposed system will predict Nitrogen status in cotton plant by real-time leave image from cotton field and display results in minimum time as compared to previous manual methods and will also make decisions how much more nitrogen will be required for its sustainable growth, estimates plant stress level due to nitrogen deficiency due to which use of fertilizer will be efficient that produce more cotton yields. This system leads to minimize costs and maximize yields.

Technical Details of Final Deliverable

• The final product we will develop both application desktop and android based they are accessible for farmers community, stakeholders as well as students, researchers.
• This system facilitates farmers to monitor cotton plant health status at different stages from sowing to picking and predict amount of nitrogen present in plant (Efficiency, Deficiency or Balanced).
• The final gadget will have a color sensor as well as a camera. Images are captured by the camera, and the RGB color value is calculated by image processing. TCS-3200 sensor detects color and calculates RGB; both findings will be merged, and a recommendation will be made.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Agriculture

Other Industries

IT

Core Technology

Internet of Things (IoT)

Other Technologies

Artificial Intelligence(AI)

Sustainable Development Goals

Zero Hunger, Reduced Inequality

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com