This paper presents a novel approach for contents-based medication image retrieval from a drug image database based on the shape and color information of the drug pill images. The shape information of the drug image represented as a signature and HSV color model from the images are used to distinguish a medication from a set of medication images. To retrieve an image from a large image database, we set up a hybrid approach which classifies the medications by shape of the images first and identifies them by color matching from the classified images. As a shape descriptor we use the signature of the medication which is unique according to the boundary of the medication. HS color components are used for retrieving a finally-matched image from a candidate shape class. The proposed system is designed and developed for seniors and people who don't have knowledge about medication to help browsing medication image in a feasible fashion.
[1]
Linda G. Shapiro,et al.
Computer Vision
,
2001
.
[2]
Ming-Kuei Hu,et al.
Visual pattern recognition by moment invariants
,
1962,
IRE Trans. Inf. Theory.
[3]
Soo-Chang Pei,et al.
Image normalization for pattern recognition
,
1995,
Image Vis. Comput..
[4]
Azriel Rosenfeld,et al.
Computer Vision
,
1988,
Adv. Comput..
[5]
Zeno Geradts,et al.
Evaluation of contents-based image retrieval methods for a database of logos on drug tablets
,
2001,
SPIE Optics East.
[6]
Elizabeth Sadler,et al.
Inclusive interface design for seniors: Image-browsing for a health information context
,
2007
.
[7]
James Ze Wang,et al.
Image retrieval: Ideas, influences, and trends of the new age
,
2008,
CSUR.