https://github.com/rifaamrilsahputra/ocr-alphanumeric-cnn
Convolutional Neural Network (CNN) for Optical Character Recognition (OCR) on alphanumeric characters (0–9 and A–Z) using the Kaggle Standard OCR Dataset.
https://github.com/rifaamrilsahputra/ocr-alphanumeric-cnn
alphanumeric character-recognition cnn deep-learning keras ocr tensorflow
Last synced: 6 months ago
JSON representation
Convolutional Neural Network (CNN) for Optical Character Recognition (OCR) on alphanumeric characters (0–9 and A–Z) using the Kaggle Standard OCR Dataset.
- Host: GitHub
- URL: https://github.com/rifaamrilsahputra/ocr-alphanumeric-cnn
- Owner: RifaAmrilSahputra
- Created: 2025-10-02T09:45:24.000Z (7 months ago)
- Default Branch: main
- Last Pushed: 2025-10-02T10:02:36.000Z (7 months ago)
- Last Synced: 2025-10-02T11:37:30.330Z (7 months ago)
- Topics: alphanumeric, character-recognition, cnn, deep-learning, keras, ocr, tensorflow
- Language: Jupyter Notebook
- Homepage:
- Size: 353 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# OCR-Alphanumeric-CNN (A–Z, 0–9)
This project builds a **Convolutional Neural Network (CNN)** model to classify handwritten characters consisting of **letters (A–Z)** and **digits (0–9)**. The dataset used is **[Standard OCR Dataset](https://www.kaggle.com/datasets/preatcher/standard-ocr-dataset/)**.
---
## ⚙️ Libraries
- TensorFlow / Keras
- NumPy
- Matplotlib
- scikit-learn
---
## 🏗️ Model Architecture
The model is based on a simple **CNN**, including:
- Conv2D + MaxPooling layers (feature extraction)
- Flatten + Dense layers (classification)
- Softmax output with **36 classes (0–9, A–Z)**
---
## 🚀 Results
- Training accuracy: **~97%**
- Validation accuracy: **~95%**
- Test accuracy: **~94%**
The model performs well in recognizing handwritten letters and digits, with minor misclassifications in similar-looking characters (e.g., `O` vs `0`).
---
## 📌 Notes
- Input images are **grayscale** with size **64x64**.