DocumentCode :
2560093
Title :
Dedicated hardware for parallel extraction of local binary pattern feature vectors
Author :
Laiho, Mika ; Lahdenoja, Olli ; Paasio, Ari
Author_Institution :
Lab. of Microelectron., Turku Univ., Finland
fYear :
2005
fDate :
28-30 May 2005
Firstpage :
27
Lastpage :
30
Abstract :
In this paper we propose a dedicated hardware for extraction of local binary pattern (LBP) feature vectors. The LBP method transforms local features of image data into binary micro-patterns that represent local and global features of the image. The LBP method can be used in applications such as texture classification, moving object detection and face detection and recognition. The hardware proposed in this paper has a massively parallel architecture in order to speed up the LBP feature extraction in real-time applications. The image data is pre-processed using an analog comparison method. Therefore, simulations are performed to find out how mismatch affects the performance.
Keywords :
feature extraction; image processing equipment; parallel architectures; real-time systems; analog comparison; dedicated hardware; local binary pattern feature vectors; parallel LBP feature extraction; parallel architecture; real-time applications; Data mining; Face detection; Face recognition; Feature extraction; Gray-scale; Hardware; Histograms; Linear discriminant analysis; Microelectronics; Principal component analysis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Cellular Neural Networks and Their Applications, 2005 9th International Workshop on
Print_ISBN :
0-7803-9185-3
Type :
conf
DOI :
10.1109/CNNA.2005.1543152
Filename :
1543152
Link To Document :
بازگشت