This project implements a Convolutional Neural Network (CNN) for image classification. The model is designed to classify images of mathematical symbols, whether they are sourced from a local directory, from the web, or hand-drawn in a paint application.
Convolutional Neural Networks (CNNs) have revolutionized image classification tasks by automatically learning features from input images. This project leverages the power of CNNs to accurately classify mathematical symbols with an impressive accuracy rate of 98%.
- Image Classification: Utilizes a CNN to classify images of mathematical symbols.
- Source Flexibility: Supports classification of symbols from local directories, web images, and hand-drawn symbols.
- High Accuracy: Achieves a classification accuracy of 98%.
-
Dataset Preparation: Organize your dataset of mathematical symbols. Ensure a diverse range of symbols are included for training.
-
Model Training: Train the CNN model using the provided dataset. Fine-tune hyperparameters to optimize performance.
-
Inference: Use the trained model to classify images of mathematical symbols. Input images can be sourced from local directories, web URLs, or from hand-drawn symbols.
- Python 3.x
- TensorFlow
- Keras
- NumPy
- Matplotlib (for visualization)
- OpenCV (for image processing)
-
Clone this repository:
git clone https://github.com/Namans12/CNN-Image-Classification-Model.git
-
Install dependencies:
pip install -r requirements.txt
-
Train the model: Train the model using your dataset or use the pre-trained weights provided.
-
Start classifying images: Use the trained model to classify images of mathematical symbols!
- This project was inspired by the advancements in deep learning and computer vision.
- Special thanks to the open-source community for providing valuable resources and libraries.