/Hindi-Alphabets-Recognition

Hindi Alphabets Recognition

Primary LanguagePython

Devanagiri Recognition

This code helps you classify different alphabets of hindi language Devanagari Characters using convolutional network.

Generated Data Animation

Code Requirements

• Need some Pre-installed Packages like

  1. Anaconda Distribution- Python 3.7

Link - https://www.anaconda.com/distribution/

  1. OpenCV – it should be installed via anaconda prompt after installation of Anaconda Distribution

Link - https://anaconda.org/conda-forge/opencv

  1. WebCam

You can install Conda for python which resolves all the dependencies for machine learning. This code is checked on :

Python version = 3.5.6
keras version = 2.2.0
Tensorflow version = 1.10.0

Description

This code successfully recognizes hindi characters.

Technique Used

I have used convolutional neural networks. I am using Tensorflow as the framework and Keras API for providing a high level of abstraction.

Architecture

CONV2D --> MAXPOOL --> CONV2D --> MAXPOOL -->FC -->Softmax--> Classification

Some additional points

  1. You can go for additional conv layers.
  2. Add regularization to prevent overfitting.
  3. You can add additional images to the training set for increasing the accuracy.

Python Implementation

  1. Dataset- DHCD (Devnagari Character Dataset) data.csv
  2. Images of size 32 X 32
  3. Convolutional Network Support added.

Train Accuracy ~ 94%

Test Accuracy ~ 93%

Execution for writing through webcam

To run the code, type `python app.py` 
python app.py