Optimized GPU implementation of JPEG 2000 for satellite image decompression
| Autor: | Ahmet Murat Ozbayoglu, Dervis Utku Ufuk, Alptekin Temizel |
|---|---|
| Přispěvatelé: | TOBB ETU, Faculty of Engineering, Department of Computer Engineering, TOBB ETÜ, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümü, Özbayoğlu, Ahmet Murat, H-2328-2011 |
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
020203 distributed computing
Image compression Computer science Remote sensing application Field programmable gate arrays (FPGA) Image coding 02 engineering and technology computer.file_format Instruction set ComputingMilieux_GENERAL Computer engineering Kernel (image processing) Asynchronous communication JPEG 2000 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing computer Decoding methods Transform coding |
| Zdroj: | CSE |
| Popis: | 21st IEEE International Conference on Computational Science and Engineering (2018 : Romania) JPEG 2000 is a powerful yet computationally complex image compression algorithm which is widely used in remote sensing applications. In this work, we focus on speeding-up the decompression algorithm by applying various GPU optimization techniques. We have conducted numerous experiments on high-resolution satellite images in two operational modes; a syn-chronous mode and an asynchronous batch mode. We share our experiment results and make performance evaluations regarding each operational mode and optimization method separately. Finally we propose an optimized GPU architecture for satellite image decompression and compare the achieved performance with a multi-threaded CPU architecture. © 2018 IEEE. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|