This code helps you recognize Handwritten English Alphabets using Convnets with an accuracy above 99%
Handwritten A-Z Alphabet Dataset from Kaggle was used to train the model. It was further used to develop a program to recognize Handwritten Text on paper and is totally suitable for real-life examples. The dataset had 3,72,450 (28x28 px) images with 26 Labels (A-Z). Out of these 0.1% was used for validation and testing each.
Keras API with Tensorflow in the backend was used to build the Deep Neural Network. OpenCV was used to detect multiple alphabets in a single image using Contours.
Training Accuracy: 99.02%
Validation Accuracy: 99.19%
Test Accuracy: 99.44%
-
Jupyter Notebook required
-
Python Libraries
- Imutils
- Tensorflow
- Keras
- Numpy
- cv2
- os
- Scikit-Learn
- Matplotlib
-
Download Jupyter Notebook
-
No further installation
There are 4 phases of this program, run each of them.
- Train a Deep Convolutional Neural Network to recognize Hand-written Alphabets A-Z (Capital Letters).
- Take a Test image with capital alphabets written on a White plain sheet.
- Localize and Detect multiple Alphabets in the in the Binary Threshold image using Contours Detection method of OpenCV
- Classify the selected Contours using the Trained DCNN model
Installing the libraries beforehand will solve most issues
Contributors names and contact info ex. @priyanshkedia04
- 0.1
- Initial Release
GNU General Public License v3.0