Plants Analysis

The proposed system will have three sub systems, for farmers, experts and admin. All the system users can use the system once they are logged in. Farmers would be able to scan the leaf image and get insights in return. The experts can provide custom feedback to the farmers and both can enhance the m

Project Title

Plants Analysis

Project Area of Specialization

Artificial Intelligence

Project Summary

The proposed system will have three sub systems, for farmers, experts and admin. All the system users can use the system once they are logged in. Farmers would be able to scan the leaf image and get insights in return. The experts can provide custom feedback to the farmers and both can enhance the machine learning model by adding new images of leaves to the current dataset. The admin can retrain the Artificial Neural Networks model once the dataset is updated by the inputs of the farmers and experts.

Flow of our proposed system is that firstly by the use of mobile device, the farmers will scan the image of leaf and send it to the server for analysis. The server will process the leaf image by using image processing techniques to extract the features and then pass it to the Artificial Neural Networks model. The Artificial Neural Networks model is trained by the annotated dataset that includes images of healthy and diseased leaves of different plants which helps the model to classify the input image into two categories: healthy or diseased.

 The reinforcement learning technique will be used by the model for classification. The system will provide the feedback to the farmers once the input is classified. If the model is not able to classify the image it will be then sent to the experts for the analysis. The experts will manually analyze the image and will provide the custom feedback to the farmers. The experts can also enhance the model by adding that feedback to the current dataset so that the model can provide the insights automatically if same type of image needs to classify in future. Admin will alert the farmers of the region where a particular disease is identified many times so that they can immediately take some precautionay measures to save the crops before it is affected by that disease. 

Project Objectives

Farmers spend billions of dollars on disease management, often without adequate technical support, resulting in poor disease control, pollution and harmful results. In addition, plant disease can devastate natural ecosystems, compounding environmental problems caused by habitat loss and poor land management. So, the purpose of building this system is to help farmers improve their crops yield before they are wasted due to plant disease. This project is based on machine learning techniques to train the model. Model will be trained using a set of healthy and diseased images of plant’s leaves. 

 The proposed system will help farmers to automatically analyze their plants and provide feedback by just scanning the image of the plant’s leaf. With the feedback provided, the farmers will get to know whether their crops will be dead or catch any disease in near future and if that is the case then they could take some precautionary measures to rescue them.

 There are some systems available to the general public which helps them to get feedback about their plants but they do not provide interfaces in different languages so that the farmers can easily use the system in their native language. However, our proposed system will provide the interface for multiple languages substantially Urdu.

Project Implementation Method

Our proposed  system will have a mobile app through which the farmers will take the plant’s leaf image and send to  server for analysis. The trained machine learning model in the server will be used to process the farmers request and send the insights back to the farmers that includes disease details, if exist, and precautionary measures to get rid of that disease. The trained machine learning model will be enhanced by the farmers and the experts input of healthier and diseased plants images. Farmers will also be able to contact the field experts if they need specialized reviews about their crops. The system will also include a catalogue which will hold the plants information and their relevant diseases. If several symptoms of the same disease is found in different parts of the region than the system will automatically alert the farmers in that region so that they can save their crops before the same disease also hit their crops. By using this system, farmers would be able to see the history of analyzed plants, search catalogue for plants information, and experts will be able to enhance the system too.   

We will use incremental methodology for this project. As our system is divided into different modules according to functionality, incremental methodology allow us to develop and test incrementally and to decompose the required product into several components depending upon functionality, each of which is designed and built separately. Since we are developing a project which uses artificial intelligence concepts (Machine and Deep Learning). So for the design part, we will use Object Oriented design methodology. This will encourage encapsulation and maintenance of the code will be easier. 

Benefits of the Project

Following are some of the benefits of the proposed system:

1. Using this app, the user can take the picture of plant’s leaf and send it to server for analysis. The server will send back the insights to the user that includes identified diseases as well as precautionary measures that help to rescue plants before the spread of disease.

2. The user only needs to take the picture using mobile and no more manual effort will be needed to detect the diseases. It will automatically be done by Artificial Intelligence models.

3. The disease detected in some specific area will be alerted to all the farmers of that area.

 4. The application will also have expert opinion if the artificial intelligence model is unable to detect the disease automatically.

5. The AI model used in application is a learning model which will be enhanced by the input provided by experts and farmers.

6. For the ease of farmers, the catalogue in application will contain all the diseases and their relevant symptoms as well as possible remedies. 

Technical Details of Final Deliverable

The source code of the whole project will be provided, that will contain a machine learning model (which will be trained with huge plants dataset), admin web application, experts web application, farmers mobile app.

The whole system will be deployed in a server where apps and model will be hosted. The mobile app will access the catalogue, machine learning model and other services through the provided restful api. All the data will be hosted in a central NoSQL database that is MongoDB and all apps will connect to it securely. The communication between the server and apps will be encrypted by using SSL over HTTP requests, which allows secure data transfer.

Final Deliverable of the Project

Software System

Type of Industry

Agriculture

Technologies

Artificial Intelligence(AI)

Sustainable Development Goals

Good Health and Well-Being for People

Required Resources

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