Clothing Identification Based on Edge Information

Author

Kim, Ku-Jin ; Park, Sun-Mi ; Choi, Yoo-Joo

Author_Institution

Dept. of Comput. Eng., Kyungpook Nat. Univ., Daegu

fYear

2008

fDate

9-12 Dec. 2008

Firstpage

876

Lastpage

880

Abstract

In this paper, we suggest a method to identify the clothing as one approach for identifying a person in the grayscale image captured by a CCTV camera. We use the edge information of the clothing area for extracting the textural feature of the clothing. The edge filter is applied to the clothing area, and the histograms on the edge values are constructed. The feature vector is composed from the histograms. We used Euclidean distance as a similarity measure between feature vectors, and successfully identified the clothing with 77.8% of success rate.

Keywords

closed circuit television; clothing; edge detection; feature extraction; image colour analysis; image texture; vectors; video surveillance; CCTV camera; Euclidean distance; clothing identification; edge filter; feature vector; grayscale image; histograms; similarity measurement; textural feature extraction; Cameras; Clothing; Color; Elevators; Filters; Gray-scale; Histograms; Image analysis; Pixel; Surveillance; cloth identification; color analysis; texture analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Asia-Pacific Services Computing Conference, 2008. APSCC '08. IEEE

Conference_Location

Yilan

Print_ISBN

978-0-7695-3473-2

Electronic_ISBN

978-0-7695-3473-2

Type

conf

DOI

10.1109/APSCC.2008.205

Filename

4780786