Skip to content

This repository is part of a research study in which, A convolution neural network(CNN) is developed from scratch for the classification of Indian coins by their denomination. The final train accuracy reached is 91.38% and the best validation accuracy is 90.62%

Notifications You must be signed in to change notification settings

ITrustNumbers/Indian_Coin_Classification_Using_CNN

Repository files navigation

A Convolution Neural Network for Classification of Indian Coins by Denomination

DOI Python 3.7 Tensforflow 2.7.0 Val_Acc License

The code is this repository is part of a research paper titled "A Novel Convolution Neural Network for Classification of Indian Coins by Denomination". The article is under the review process as of December 2021 but a 'Preprint' of the article can be found here Preprint
DOI :

About The Data:

Indian Coin Denomination Dataset(ICDD): (IEEE DataPort)

The data used in this study is collected by the authors and stored using the follow filling structure:

dataset/
     test_set/
            five/
            one/
            ten/
            twenty/
            two/
      training_set/
            five/
            one/
            ten/
            twenty/
            two/

The data contains images of One (1), Two (2), Five (5), Ten (10), and Twenty (20) Rupee coins in Indian denomination and their variations that are in circulation as of December 2021. The dataset only contains images of one side of each coin (Tail side) which contains the denomination value.

In total the dataset contains 900 images out of which there are 210 images of one rupee class, 120 images of two rupee class, 270 images of five rupee class, 180 images of ten rupee class, and another 120 images of twenty rupee class.

For more information on the data refer to the documentation available on IEEE DataPort here Indian Coin Denomination Dataset(ICDD)

Citation:

For Citing The Article use bellow citation(APA Format):

Authors:
Yash Chauhan, Prateek Singh
    
Citation: 
Chauhan, Y., & Singh, P. (2021). A Novel Convolutional Neural Network for Classifying Indian Coins by Denomination. https://doi.org/10.31224/osf.io/znxrg

DOI: 
10.31224/osf.io/znxrg

For Citation in other format than APA go Here

For Citing the Dataset use bellow citation(APA Format):

Authors:
Yash Chauhan, Prateek Singh
    
Citation:
Yash Chauhan, Prateek Singh. (2021). Indian Coin Denomination Dataset (ICDD). IEEE Dataport. https://dx.doi.org/10.21227/t63t-5j24
    
DOI:
10.21227/t63t-5j24

For Citation in other format than APA go Here

About

This repository is part of a research study in which, A convolution neural network(CNN) is developed from scratch for the classification of Indian coins by their denomination. The final train accuracy reached is 91.38% and the best validation accuracy is 90.62%

Topics

Resources

Stars

Watchers

Forks

Languages