Paper Info
Title
TreeBASIS feature descriptor and its hardware implementation
Abstract
This paper presents a novel feature descriptor called TreeBASIS that provides improvements in descriptor size, computation time, matching speed, and accuracy. This new descriptor uses a binary vocabulary tree that is computed using basis dictionary images and a test set of feature region images. To facilitate real-time implementation, a feature region image is binary quantized and the resulting quantized vector is passed into the BASIS vocabulary tree. A Hamming distance is then computed between the feature region image and the effectively descriptive basis dictionary image at a node to determine the branch taken and the path the feature region image takes is saved as a descriptor. The TreeBASIS feature descriptor is an excellent candidate for hardware implementation because of its reduced descriptor size and the fact that descriptors can be created and features matched without the use of floating point operations. The TreeBASIS descriptor is more computationally and space efficient than other descriptors such as BASIS, SIFT, and SURF. Moreover, it can be computed entirely in hardware without the support of a CPU for additional software-based computations. Experimental results and a hardware implementation show that the TreeBASIS descriptor compares well with other descriptors for frame-to-frame homography computation while requiring fewer hardware resources.
Year
DOI
Venue
2014
10.1155/2014/606210
International Journal of Reconfigurable Computing
Field
DocType
Volume
Scale-invariant feature transform,Floating point,Computer science,Local binary patterns,Software,Homography,Artificial intelligence,Computer hardware,Computer vision,Pattern recognition,GLOH,Hamming distance,Test set
Journal
2014
Issue
Citations 
PageRank 
1
1
0.37
References 
Authors
24
5
Name
Order
Citations
PageRank
Spencer G. Fowers1384.74
Alok Desai2204.17
Dah-Jye Lee342242.05
Dan Ventura4314.39
James K. Archibald5632161.01