Napredna pretraga

Pregled bibliografske jedinice broj: 527903

Acceleration of Ceramic Tiles Machine Vision Quality Control Algorithm Using CUDA


Hocenski, Željko; Matić, Tomislav
Acceleration of Ceramic Tiles Machine Vision Quality Control Algorithm Using CUDA // Proceedings of SICE Annual Conference 2010 / Osuka, Coichi ; Li, Tzuu-Hseng S. (ur.).
Taipei, Taiwan: SICE 2010, 2010. str. 2170-2174 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)


Naslov
Acceleration of Ceramic Tiles Machine Vision Quality Control Algorithm Using CUDA

Autori
Hocenski, Željko ; Matić, Tomislav

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings of SICE Annual Conference 2010 / Osuka, Coichi ; Li, Tzuu-Hseng S. - Taipei, Taiwan : SICE 2010, 2010, 2170-2174

ISBN
978-4-907764-35-7

Skup
SICE Annual Conference 2010

Mjesto i datum
Taipei, Taiwan, 18-21.08.2010

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Quality control ; CUDA ; algorithm ; ceramic tiles ; image processing ; parallel processing

Sažetak
Speedup of the processing algorithm for automated visual quality control of ceramic tiles is presented in this paper. The quality classification of ceramic tiles depends on dimensions and surface features. Presented machine vision quality control algorithm analyzes surface features looking for defects and returns number of found defects. Machine vision system for ceramic tiles is a real time system with hard deadline, processing time limited by production line speed. Image processing is time consuming and it could become too slow. The problem increases with tile dimension, color and relief. One of the solutions to decrease the processing time is the usage of parallel processing. The main idea of this paper is the implementation of CUDA (Compute Unified Device Architecture) technology for image processing in visual control of ceramic tiles. It is new technology of general-purpose computing on the GPU (Graphics Processing Unit) that enables easier development of GPU programs. For the purpose of this paper a simple algorithm was written in C programming language that was implemented in the CPU and the GPU. The processing time results and the speedup are compared between the CPU and the GPU for some examples with several dimensions and surface design.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo



POVEZANOST RADA


Projekt / tema
165-0361621-2000 - Distribuirano računalno upravljanje u transportu i industrijskim pogonima (Željko Hocenski, )

Ustanove
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek