Introduction

This project use feature extractor and classification algorithm to detect 0 - 11 digit of Hoda dataset.

Dataset

• HODA Farsi Digit Dataset : http://farsiocr.ir

Classification Algorithm

• KNN
• Bayes
• MLP
• RBF
• AdaBoost
• GSD
• Parzen
• LinearSVC

Feature Extractor Algorithm

• HOG
• SVD
• PCA

Architecture

Code Requirements

This code is written in python. To use it you will need:

• python3
• matplotlib
• sklearn
• numpy
• cv2

Usage

Run python PHDRI.py

Result

Algorithm	Feature-Extrector	Accuracy
KNN	HOG	91%
KNN	SVD	94%
KNN	PCA	85%
MLP	HOG	93%
MLP	SVD	57%
MLP	PCA	90%
Parzen	HOG	91%
Parzen	SVD	70%
Parzen	PCA	83.2%
Bayes	HOG	91%
Bayes	SVD	75%
Bayes	PCA	85%
AdaBoost	HOG	48%
AdaBoost	SVD	50%
AdaBoost	PCA	30%
LinearSVC	HOG	85%
LinearSVC	SVD	70%
LinearSVC	PCA	60%
RBF	HOG	88%
RBF	SVD	83%
RBF	PCA	80%