Home // MMEDIA 2011, The Third International Conferences on Advances in Multimedia // View article

Accelerating Image Processing in Flash using SIMD Standard Operations

Authors:
Chamira Perera
Daniel Shapiro
Jonathan Parri
Miodrag Bolic
Voicu Groza

Keywords: SIMD; image processing; native code interface; image processing acceleration; virtual machine; Flash

Abstract:
Flash applications have played an integral role in shaping the interactivity of the Internet. Desktop Flash applications feature vector-based processing such as image and video processing to enhance the user experience. In response to these needs, Adobe has added graphics card based acceleration for vector processing in Flash applications starting with Flash Player 10. This solution is limited to computer systems that have the proper graphics card. In this paper, we investigate the possibility of making explicit use of Single Instruction Multiple Data instructions, specifically SSE in the Intel x86-64 platforms, to accelerate vector operations in a Flash application. We also discuss certain limitations of the Flash virtual machine. The data reveals that a 90-92% speedup can be achieved by using SSE instructions to accelerate the alpha blending image processing algorithm in a Flash application. The SSE instructions are accessed by providing a standardized limited native interface to the Flash application.

Pages: 118 to 123

Copyright: Copyright (c) IARIA, 2011

Publication date: April 17, 2011

Published in: conference

ISSN: 2308-4448

ISBN: 978-1-61208-129-8

Location: Budapest, Hungary

Dates: from April 17, 2011 to April 22, 2011