Detection of pests in agriculture using Machine Learning

Hey Guys! I’ve started off a project using Machine learning and Neural networks. Its a pretty simple project with the pre-requisites being proficient in the basics of ML and DL. Feel free to join in this quest of completing the project :slight_smile:

Project description: We are going to train a model to identify healthy plant leaves and infected plant leaves by providing a set of data for training and testing.

How will we proceed:

  1. Gathering data: We’ll start off with collecting the data of infected plant leaves and healthy plant leaves.
    I already got about 1.5 Gigs of the data but it’s unlabelled for now.

  2. Coding part: We’ll use Python to train the model. We will be importing Sci-kit, Tensorflow libraries and it will be a pretty easy task after importing those libraries. I suggest you to refer a course in AI with Python.

  3. Testing and training: We can divide the data we’ve collected into two sets for testing and training. We can again use Python to test and train our model.

  4. Output: We will develop an AI model such that when we show it a leaf, it will identify whether the leaf is infected or not.

Extensions: The project that I’ve explained above is just about creating an AI model.

There are a lot of possible extensions such as developing a front end application for our AI and so many more possibilities. We can discuss them too

Any suggestions are accepted :slight_smile:


@Dantusaikamal The idea is pretty good and feasible. But, can you elaborate your idea somewhat in detail ? Who gets more benifitted and how would the get benefitted?

1 Like

I think your project idea is similar to “Automatic leaf infection identification” . Did you check this out, if not use this as reference.

Yes sure. I was going to edit the thread to further elaborate the idea. So basically we take images of two sets. One is healthy plant leaves and the other is infected plant leaves. We are going to label the data and then use python to train the ML model to identify whether a plant is healthy or infected

1 Like

Thank you, Akshith. I searched for existing ones but couldn’t find a reliable one. I’ll go through the repository. It’ll definitely be helpful :slightly_smiling_face:


Thank you Bhanu. This is completely unrelated to what we’ll be doing in this project but I wanted to develop an Agrodrone which will use it’s sensory data to identify if a farmer’s crop is infected with pests. Agrodrones already exist but are really costly and not user friendly for farmers. I want to develop a drone which can help in irrigation, identification of pest, and so much more at the cheapest cost possible. It’s not going to be an easy task and involves a lot of questions but it’s definitely possible :slight_smile:


I have a small question…Are you going to detect the type of pest here or are you going to classify whether the plant leaves are healthy or not?

1 Like

Let’s start with identifying whether the leaves are healthy or infected. We can later on collect data on different types of pest and train the model to classify the type of pest.

1 Like

So there is a project called as the pest detection project it is almost the same idea, you can choose that project. The main idea of the project Pest Detection is that it takes samples and images of leaves and tells what kind of pest it is.

1 Like