Accurate system for automatic pill recognition using imprint information

Jiye Yu, Zhiyuan Chen, Seiichiro Kamata, Jie Yang

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

With rapidly advancing of contemporary medicine, it is necessary to help people identify various kinds of pills to prevent the adverse pill events. In this study, a high-accuracy automatic pill recognition system is proposed for accurate and automatic pill recognition. As pill imprint is main distinction between different pills, this system proposes algorithms on both imprint extraction and description parts to make use of imprint information. First, proposed modified stroke width transform is adopted to extract the imprint by detecting coherent strokes of imprint on the pill. Moreover, image segmentation by Loopy belief propagation is also added on printed imprint pills to solve the incoherent and coarse stroke problem. Second, a new descriptor named two-step sampling distance sets is proposed for accurate imprint description and successfully cut down the noise on extracted imprint. This strategy is based on the imprint partition - partitions the imprint on the basis of separated strokes, fragments and noise points. Recognition experiments are applied on extensive databases and result shows 90.46% rank-1 matching accuracy and 97.16% on top five ranks when classifying 12 500 query pill images into 2500 categories.

Original languageEnglish
Pages (from-to)1039-1047
Number of pages9
JournalIET Image Processing
Volume9
Issue number12
DOIs
Publication statusPublished - 2015 Dec 1

Fingerprint

Image segmentation
Medicine
Sampling
Experiments

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering
  • Software
  • Computer Vision and Pattern Recognition

Cite this

Accurate system for automatic pill recognition using imprint information. / Yu, Jiye; Chen, Zhiyuan; Kamata, Seiichiro; Yang, Jie.

In: IET Image Processing, Vol. 9, No. 12, 01.12.2015, p. 1039-1047.

Research output: Contribution to journalArticle

Yu, Jiye ; Chen, Zhiyuan ; Kamata, Seiichiro ; Yang, Jie. / Accurate system for automatic pill recognition using imprint information. In: IET Image Processing. 2015 ; Vol. 9, No. 12. pp. 1039-1047.
@article{77f4fd410d1343029ba388ccf11e57e0,
title = "Accurate system for automatic pill recognition using imprint information",
abstract = "With rapidly advancing of contemporary medicine, it is necessary to help people identify various kinds of pills to prevent the adverse pill events. In this study, a high-accuracy automatic pill recognition system is proposed for accurate and automatic pill recognition. As pill imprint is main distinction between different pills, this system proposes algorithms on both imprint extraction and description parts to make use of imprint information. First, proposed modified stroke width transform is adopted to extract the imprint by detecting coherent strokes of imprint on the pill. Moreover, image segmentation by Loopy belief propagation is also added on printed imprint pills to solve the incoherent and coarse stroke problem. Second, a new descriptor named two-step sampling distance sets is proposed for accurate imprint description and successfully cut down the noise on extracted imprint. This strategy is based on the imprint partition - partitions the imprint on the basis of separated strokes, fragments and noise points. Recognition experiments are applied on extensive databases and result shows 90.46{\%} rank-1 matching accuracy and 97.16{\%} on top five ranks when classifying 12 500 query pill images into 2500 categories.",
author = "Jiye Yu and Zhiyuan Chen and Seiichiro Kamata and Jie Yang",
year = "2015",
month = "12",
day = "1",
doi = "10.1049/iet-ipr.2014.1007",
language = "English",
volume = "9",
pages = "1039--1047",
journal = "IET Image Processing",
issn = "1751-9659",
publisher = "Institution of Engineering and Technology",
number = "12",

}

TY - JOUR

T1 - Accurate system for automatic pill recognition using imprint information

AU - Yu, Jiye

AU - Chen, Zhiyuan

AU - Kamata, Seiichiro

AU - Yang, Jie

PY - 2015/12/1

Y1 - 2015/12/1

N2 - With rapidly advancing of contemporary medicine, it is necessary to help people identify various kinds of pills to prevent the adverse pill events. In this study, a high-accuracy automatic pill recognition system is proposed for accurate and automatic pill recognition. As pill imprint is main distinction between different pills, this system proposes algorithms on both imprint extraction and description parts to make use of imprint information. First, proposed modified stroke width transform is adopted to extract the imprint by detecting coherent strokes of imprint on the pill. Moreover, image segmentation by Loopy belief propagation is also added on printed imprint pills to solve the incoherent and coarse stroke problem. Second, a new descriptor named two-step sampling distance sets is proposed for accurate imprint description and successfully cut down the noise on extracted imprint. This strategy is based on the imprint partition - partitions the imprint on the basis of separated strokes, fragments and noise points. Recognition experiments are applied on extensive databases and result shows 90.46% rank-1 matching accuracy and 97.16% on top five ranks when classifying 12 500 query pill images into 2500 categories.

AB - With rapidly advancing of contemporary medicine, it is necessary to help people identify various kinds of pills to prevent the adverse pill events. In this study, a high-accuracy automatic pill recognition system is proposed for accurate and automatic pill recognition. As pill imprint is main distinction between different pills, this system proposes algorithms on both imprint extraction and description parts to make use of imprint information. First, proposed modified stroke width transform is adopted to extract the imprint by detecting coherent strokes of imprint on the pill. Moreover, image segmentation by Loopy belief propagation is also added on printed imprint pills to solve the incoherent and coarse stroke problem. Second, a new descriptor named two-step sampling distance sets is proposed for accurate imprint description and successfully cut down the noise on extracted imprint. This strategy is based on the imprint partition - partitions the imprint on the basis of separated strokes, fragments and noise points. Recognition experiments are applied on extensive databases and result shows 90.46% rank-1 matching accuracy and 97.16% on top five ranks when classifying 12 500 query pill images into 2500 categories.

UR - http://www.scopus.com/inward/record.url?scp=84948968396&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84948968396&partnerID=8YFLogxK

U2 - 10.1049/iet-ipr.2014.1007

DO - 10.1049/iet-ipr.2014.1007

M3 - Article

VL - 9

SP - 1039

EP - 1047

JO - IET Image Processing

JF - IET Image Processing

SN - 1751-9659

IS - 12

ER -