Related Experiment Video
Updated: Dec 5, 2025
05:41
A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
Published on: February 6, 2020
9.7K
Recognition of Pashto Handwritten Characters Based on Deep Learning.
Muhammad Sadiq Amin1, Siddiqui Muhammad Yasir1, Hyunsik Ahn1
1Department of Robot System Engineering, Tongmyong University, Busan 48520, Korea.
Sensors (Basel, Switzerland)
|October 21, 2020
Summary
This study introduces a novel convolutional neural network (CNN) for Pashto handwritten character recognition (PHCR), achieving 99.64% accuracy. The developed model and dataset offer a significant advancement for automated Pashto character recognition systems.
Area of Science:
- Computer Science
- Artificial Intelligence
- Machine Learning
Background:
- Handwritten character recognition is crucial for automation in various fields.
- Pashto handwritten character recognition (PHCR) faces challenges due to complex characters and limited research.
- Existing object recognition technologies are insufficient for accurate PHCR.
Purpose of the Study:
- To propose a novel Convolutional Neural Network (CNN) model for Pashto handwritten character recognition (PHCR) in unrestricted environments.
- To develop and evaluate a new dataset for Pashto handwritten characters.
- To establish a benchmark for PHCR performance using state-of-the-art deep learning models.
Main Methods:
- Construction of a novel Pashto handwritten character dataset named "Poha" with 44 characters.
- Application of deep fusion image processing and noise reduction techniques for text optimization.
- Development and optimization of a CNN model, comparing its performance against popular benchmark CNN models.
Main Results:
- The proposed CNN model achieved a test accuracy of 99.64% for Pashto handwritten character recognition.
- The optimized CNN model demonstrated superior performance compared to common deep learning models on the Poha dataset.
- Experimental results confirm the effectiveness of the proposed approach for PHCR.
Conclusions:
- The developed CNN model represents a significant advancement in Pashto handwritten character recognition.
- The proposed model shows strong potential for automated PHCR applications.
- This research addresses a critical gap in the field of handwritten character recognition for the Pashto language.
