Tiling and demand driven evaluation for picture processing

Patrick Horain; Victor Dogaru

doi:10.1117/12.216369

Tiling and demand driven evaluation for picture processing

Patrick Horain
, Victor Dogaru

Telecom Paris

Research output: Contribution to journal › Conference article › peer-review

Abstract

We propose a software architecture for picture processing that allows efficient memory management when algorithms with many operators are applied to large images, and that allows automated parallelization. This architecture relies on image tiling and operators with a call back function that evaluate image tiles on demand. Several tiling strategies with and without overlapping are discussed. The complexity of this evaluation strategy is hidden to application programs. This is shown with a sample program. This architecture is well suited for neighborhood operators such as convolutions and mathematical morphology operators.

Original language	English
Pages (from-to)	87-93
Number of pages	7
Journal	Proceedings of SPIE - The International Society for Optical Engineering
Volume	2568
DOIs	https://doi.org/10.1117/12.216369
Publication status	Published - 11 Aug 1995
Event	Neural, Morphological, and Stochastic Methods in Image and Signal Processing 1995 - San Diego, United States Duration: 9 Jul 1995 → 14 Jul 1995

Keywords

Demand driven evaluation
Image processing
Memory management
Parallelism
Tiling

Access to Document

10.1117/12.216369

Cite this

@article{bfc1c26307844c478c1d363ee564aa2e,

title = "Tiling and demand driven evaluation for picture processing",

abstract = "We propose a software architecture for picture processing that allows efficient memory management when algorithms with many operators are applied to large images, and that allows automated parallelization. This architecture relies on image tiling and operators with a call back function that evaluate image tiles on demand. Several tiling strategies with and without overlapping are discussed. The complexity of this evaluation strategy is hidden to application programs. This is shown with a sample program. This architecture is well suited for neighborhood operators such as convolutions and mathematical morphology operators.",

keywords = "Demand driven evaluation, Image processing, Memory management, Parallelism, Tiling",

author = "Patrick Horain and Victor Dogaru",

note = "Publisher Copyright: {\textcopyright} 1995 SPIE.; Neural, Morphological, and Stochastic Methods in Image and Signal Processing 1995 ; Conference date: 09-07-1995 Through 14-07-1995",

year = "1995",

month = aug,

day = "11",

doi = "10.1117/12.216369",

language = "English",

volume = "2568",

pages = "87--93",

journal = "Proceedings of SPIE - The International Society for Optical Engineering",

issn = "0277-786X",

}

TY - JOUR

T1 - Tiling and demand driven evaluation for picture processing

AU - Horain, Patrick

AU - Dogaru, Victor

PY - 1995/8/11

Y1 - 1995/8/11

N2 - We propose a software architecture for picture processing that allows efficient memory management when algorithms with many operators are applied to large images, and that allows automated parallelization. This architecture relies on image tiling and operators with a call back function that evaluate image tiles on demand. Several tiling strategies with and without overlapping are discussed. The complexity of this evaluation strategy is hidden to application programs. This is shown with a sample program. This architecture is well suited for neighborhood operators such as convolutions and mathematical morphology operators.

AB - We propose a software architecture for picture processing that allows efficient memory management when algorithms with many operators are applied to large images, and that allows automated parallelization. This architecture relies on image tiling and operators with a call back function that evaluate image tiles on demand. Several tiling strategies with and without overlapping are discussed. The complexity of this evaluation strategy is hidden to application programs. This is shown with a sample program. This architecture is well suited for neighborhood operators such as convolutions and mathematical morphology operators.

KW - Demand driven evaluation

KW - Image processing

KW - Memory management

KW - Parallelism

KW - Tiling

UR - https://www.scopus.com/pages/publications/67649170270

U2 - 10.1117/12.216369

DO - 10.1117/12.216369

M3 - Conference article

AN - SCOPUS:67649170270

SN - 0277-786X

VL - 2568

SP - 87

EP - 93

JO - Proceedings of SPIE - The International Society for Optical Engineering

JF - Proceedings of SPIE - The International Society for Optical Engineering

T2 - Neural, Morphological, and Stochastic Methods in Image and Signal Processing 1995

Y2 - 9 July 1995 through 14 July 1995

ER -

Tiling and demand driven evaluation for picture processing

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this