Press/กด CtrlหรือCmd + P to print
or save as PDF

Pill Classification Based on Machine Learning

Estimated reading time: 1 min
Research Article: Pill Classification Based on Machine Learning
Author: Areerat Pathomchaiwal
Email: areerat.pat@siam.edu
Department|Faculty: Department of Printing Engineering, Faculty of Engineering, Institute of Printing Engineering, Siam University, Thailand
Published: The 6th Asian Color Association (ACA) Conference 2021 “Color & Culture” 02-03 November 2021, Yogyakarta Indonesia

Citation

Pathomchaiwal A. (2021). Pill classification based on machine learning. In The 6th Asian Color Association (ACA) Conference 2021 “Color & Culture” 02-03 November 2021, Yogyakarta Indonesia (pp.38-41). Indonesia: Department of Agro-Industrial Technology Universitas Gadjah Mada.


ABSTRACT

Pill image classification has been widely used in commercial applications. In this paper, we propose a deep learning for pill image classification. The proposed pill dataset into separate training and testing sets. We find that neural networks need significantly less training data to obtain the state-of-the-art performance than previously proposed methods. This method is based on a residual neural network for pill image classification. The results show that the proposed classification model yields an accuracy of 97.27% in classification performance.

Keywords: Pill Image, Machine Learning (ML), Residual Neural Network (ResNet), Imbalanced Data, Classification.


Pill Classification Based on Machine Learning

Faculty of Engineering, Siam University, Bangkok, Thailand