Feature based object recognition using discrete wavelet transform

Author

Elakkiya, S. ; Audithan, S.

Author_Institution

PRIST Univ., Tanjore, India

fYear

2014

fDate

8-8 July 2014

Firstpage

393

Lastpage

396

Abstract

In this paper, an approach for object recognition based on wavelet transform is presented. This approach decomposes the input image into sub-bands by using the multiresolutional analysis, Discrete Wavelet Transform (DWT). As each sub-band in the decomposed image contains useful information about the object, the mean of each sub-band is considered as features. This approach is tested on Columbia Object Image Library (COIL-100) database. All the objects in the COIL-100 database are considered for the classification based on one nearest neighbor classifier. The result shows that maximum recognition accuracy of 85.72% is achieved by the wavelet features.

Keywords

discrete wavelet transforms; feature extraction; learning (artificial intelligence); object recognition; COIL-100 database; Columbia Object Image Library database; DWT; discrete wavelet transform; feature based object recognition; image decomposition; multiresolutional analysis; nearest neighbor classifier; Accuracy; Databases; Discrete wavelet transforms; Feature extraction; Image color analysis; Object recognition; Training; Haar wavelet; object recognition; one nearest neighbor; wavelet transform;

fLanguage

English

Publisher

ieee

Conference_Titel

Current Trends in Engineering and Technology (ICCTET), 2014 2nd International Conference on

Conference_Location

Coimbatore

Print_ISBN

978-1-4799-7986-8

Type

conf

DOI

10.1109/ICCTET.2014.6966323

Filename

6966323