Pregled bibliografske jedinice broj: 398128
Optimizing Image Processing in C#
Optimizing Image Processing in C# // Proceedings Vol. III. CTS & CIS / Bogunović, Nikola ; Ribarić, Slobodan (ur.).
Opatija: Hrvatska udruga za informacijsku i komunikacijsku tehnologiju, elektroniku i mikroelektroniku - MIPRO, 2009. str. 30-34 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 398128 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Optimizing Image Processing in C#
Autori
Filko, Damir ; Antonić, Davor
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings Vol. III. CTS & CIS
/ Bogunović, Nikola ; Ribarić, Slobodan - Opatija : Hrvatska udruga za informacijsku i komunikacijsku tehnologiju, elektroniku i mikroelektroniku - MIPRO, 2009, 30-34
ISBN
978-953-233-045-8
Skup
32nd International Convention on Information and Communication Technology, Electronics and Microelectronics
Mjesto i datum
Opatija, Hrvatska, 25.05.2009. - 29.05.2009
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
image processing; memory management; parallel processing; C#
Sažetak
In modern world image analysis takes important role in all aspects of technical sciences, from product analysis on production lines to robotic vision and medical applications. Algorithm implementation in C# programming language is characterized by rapid development, resulting in highly managed and safe code. Price paid is low efficiency, which is especially significant in algorithms processing large volumes of data such as image processing algorithms. In this paper two complementary methods of increasing execution speed of such algorithms are implemented. Primary source of poor C# code performance results from the way image data is accessed from memory. Array bound checking embedded in managed code and accessing data through image objects imposes large amount of overhead. Utilizing low-level functions and procedures for accessing data could increase performance for an order of magnitude. Another way to increase execution speed is to utilize parallelism generally existing in majority of image processing algorithms. Executing parallel code on multi-core or the multiprocessor system further increases execution speed.
Izvorni jezik
Engleski