Giter Club home page Giter Club logo

satellite-image-classification's Introduction

Satellite Image Classification

This repository contains code for building a deep learning model to classify satellite images into different categories such as Cloudy, Desert, Green Area, and Water.

What It Does

The model is designed to take satellite images as input and predict the category of the image, providing insights into the type of terrain or environment captured in the satellite data.

How It Works

  1. Data Preparation:

    • The data can be obtained from Kaggle, where it's organized into folders representing different categories (e.g., Cloudy, Desert, Green Area, Water).
    • Image paths and labels are extracted from the folder structure and saved into a CSV file for easier handling.
  2. Data Preprocessing:

    • Images are resized and preprocessed using techniques like rescaling, shear, zoom, flip, and rotation to enhance the model's ability to generalize.
  3. Model Architecture:

    • A deep learning model is built using Convolutional Neural Networks (CNNs) for feature extraction.
    • The model architecture includes multiple Conv2D layers with activation functions like ReLU, MaxPooling layers for spatial downsampling, a Flatten layer to convert the 2D feature maps into a 1D vector, and Dense layers for classification.
    • Dropout is used to prevent overfitting by randomly disabling neurons during training.
  4. Training and Evaluation:

    • The model is trained using the prepared dataset with a specified number of epochs.
    • Training and validation metrics such as loss and accuracy are monitored and visualized using matplotlib.
  5. Model Deployment:

    • After training, the model is saved as a .h5 file for future use.
  6. Prediction:

    • The trained model is used to make predictions on new satellite images.
    • Users can change the image path in run_model.py to any satellite image of their choice, as long as it is the right size (255x255 pixels).

Tools and Methods Used

  • Python Libraries: pandas, os, numpy, sklearn, keras, PIL, matplotlib, tkinter
  • Data Augmentation: ImageDataGenerator for generating augmented images during training.
  • Model Building: Sequential model with Conv2D, MaxPooling2D, Flatten, Dense, and Dropout layers.
  • Model Evaluation: Metrics such as accuracy and loss are used to evaluate model performance.
  • Visualization: Matplotlib is used to visualize training and validation metrics.

Usage

  1. Clone the repository to your local machine.
  2. Install the required libraries.
  3. Obtain the dataset from Kaggle and organize it into folders as per the provided structure.
  4. Run create_model.py to train the model and save it as Model.h5.
  5. Use run_model.py to load the trained model and make predictions on new satellite images by changing the image path to your desired satellite image of the right size (255x255 pixels).

Example Result

predictionui

satellite-image-classification's People

Contributors

tawsifmayaz avatar tawsifrm avatar

Watchers

 avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.