Scalable Data-Parallel Implementations of Object Recognition Using Geometric Hashing

C. L. Wang, V. K. Prasanna, Hyong Joong Kim, A. A. Khokhar

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Object recognition involves identifying known objects in a given scene. It plays a key role in image understanding. Geometric hashing has been proposed as a technique for model-based object recognition in occluded scenes. However, parallel techniques are needed to realize real time vision systems employing geometric hashing. In this paper, we present scalable parallel algorithms for object recognition using geometric hashing. We define a realistic abstract model of CM-5 in which explicit cost is associated with data routing and synchronization. We develop a load-balancing technique that results in scalable processor-time optimal algorithms for performing a probe on this model. Given a model of CM-5 with P PNs and a set S of feature points in a scene, a probe of the recognition phase can be performed in O(|V(S)|/P) time, where V(S) is the set of votes cast by feature points in S. This algorithm is scalable in the range 1 ≤ P ≤ |V(S)| 1 3. On a mesh processor array of any size [formula] × [formula] which models MP-1, we show that a probe can be performed on O(|V(S)|/[formula]) time, log2|V(S)| ≤ P ≤ |V(S)|. These results do not assume any distributions of hash bin lengths or scene points. In earlier parallel implementations, the number of processors employed was independent of the size of the scene but depended on the size of the model database (which is usually very large). Our implementations on CM-5 and MP-1 significantly improve upon the number of processors employed and also result in superior time performance. The implementations developed in this paper require a number of processors that are independent of the size of the model database and are scalable with the machine size. Results of concurrent processing of multiple probes are also reported.

Original languageEnglish
Pages (from-to)96-109
Number of pages14
JournalJournal of Parallel and Distributed Computing
Volume21
Issue number1
DOIs
Publication statusPublished - 1994 Apr
Externally publishedYes

Fingerprint

Hashing
Object recognition
Object Recognition
Parallel Implementation
Probe
Feature Point
Model
Image Understanding
Real-time Systems
Image understanding
Vote
Vision System
Optimal Algorithm
Load Balancing
Parallel Algorithms
Bins
Parallel processing systems
Concurrent
Parallel algorithms
Routing

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Hardware and Architecture
  • Computer Science Applications

Cite this

Scalable Data-Parallel Implementations of Object Recognition Using Geometric Hashing. / Wang, C. L.; Prasanna, V. K.; Kim, Hyong Joong; Khokhar, A. A.

In: Journal of Parallel and Distributed Computing, Vol. 21, No. 1, 04.1994, p. 96-109.

Research output: Contribution to journalArticle

@article{a7cf11babd2b41449b3681b1a3f3d668,
title = "Scalable Data-Parallel Implementations of Object Recognition Using Geometric Hashing",
abstract = "Object recognition involves identifying known objects in a given scene. It plays a key role in image understanding. Geometric hashing has been proposed as a technique for model-based object recognition in occluded scenes. However, parallel techniques are needed to realize real time vision systems employing geometric hashing. In this paper, we present scalable parallel algorithms for object recognition using geometric hashing. We define a realistic abstract model of CM-5 in which explicit cost is associated with data routing and synchronization. We develop a load-balancing technique that results in scalable processor-time optimal algorithms for performing a probe on this model. Given a model of CM-5 with P PNs and a set S of feature points in a scene, a probe of the recognition phase can be performed in O(|V(S)|/P) time, where V(S) is the set of votes cast by feature points in S. This algorithm is scalable in the range 1 ≤ P ≤ |V(S)| 1 3. On a mesh processor array of any size [formula] × [formula] which models MP-1, we show that a probe can be performed on O(|V(S)|/[formula]) time, log2|V(S)| ≤ P ≤ |V(S)|. These results do not assume any distributions of hash bin lengths or scene points. In earlier parallel implementations, the number of processors employed was independent of the size of the scene but depended on the size of the model database (which is usually very large). Our implementations on CM-5 and MP-1 significantly improve upon the number of processors employed and also result in superior time performance. The implementations developed in this paper require a number of processors that are independent of the size of the model database and are scalable with the machine size. Results of concurrent processing of multiple probes are also reported.",
author = "Wang, {C. L.} and Prasanna, {V. K.} and Kim, {Hyong Joong} and Khokhar, {A. A.}",
year = "1994",
month = "4",
doi = "10.1006/jpdc.1994.1044",
language = "English",
volume = "21",
pages = "96--109",
journal = "Journal of Parallel and Distributed Computing",
issn = "0743-7315",
publisher = "Academic Press Inc.",
number = "1",

}

TY - JOUR

T1 - Scalable Data-Parallel Implementations of Object Recognition Using Geometric Hashing

AU - Wang, C. L.

AU - Prasanna, V. K.

AU - Kim, Hyong Joong

AU - Khokhar, A. A.

PY - 1994/4

Y1 - 1994/4

N2 - Object recognition involves identifying known objects in a given scene. It plays a key role in image understanding. Geometric hashing has been proposed as a technique for model-based object recognition in occluded scenes. However, parallel techniques are needed to realize real time vision systems employing geometric hashing. In this paper, we present scalable parallel algorithms for object recognition using geometric hashing. We define a realistic abstract model of CM-5 in which explicit cost is associated with data routing and synchronization. We develop a load-balancing technique that results in scalable processor-time optimal algorithms for performing a probe on this model. Given a model of CM-5 with P PNs and a set S of feature points in a scene, a probe of the recognition phase can be performed in O(|V(S)|/P) time, where V(S) is the set of votes cast by feature points in S. This algorithm is scalable in the range 1 ≤ P ≤ |V(S)| 1 3. On a mesh processor array of any size [formula] × [formula] which models MP-1, we show that a probe can be performed on O(|V(S)|/[formula]) time, log2|V(S)| ≤ P ≤ |V(S)|. These results do not assume any distributions of hash bin lengths or scene points. In earlier parallel implementations, the number of processors employed was independent of the size of the scene but depended on the size of the model database (which is usually very large). Our implementations on CM-5 and MP-1 significantly improve upon the number of processors employed and also result in superior time performance. The implementations developed in this paper require a number of processors that are independent of the size of the model database and are scalable with the machine size. Results of concurrent processing of multiple probes are also reported.

AB - Object recognition involves identifying known objects in a given scene. It plays a key role in image understanding. Geometric hashing has been proposed as a technique for model-based object recognition in occluded scenes. However, parallel techniques are needed to realize real time vision systems employing geometric hashing. In this paper, we present scalable parallel algorithms for object recognition using geometric hashing. We define a realistic abstract model of CM-5 in which explicit cost is associated with data routing and synchronization. We develop a load-balancing technique that results in scalable processor-time optimal algorithms for performing a probe on this model. Given a model of CM-5 with P PNs and a set S of feature points in a scene, a probe of the recognition phase can be performed in O(|V(S)|/P) time, where V(S) is the set of votes cast by feature points in S. This algorithm is scalable in the range 1 ≤ P ≤ |V(S)| 1 3. On a mesh processor array of any size [formula] × [formula] which models MP-1, we show that a probe can be performed on O(|V(S)|/[formula]) time, log2|V(S)| ≤ P ≤ |V(S)|. These results do not assume any distributions of hash bin lengths or scene points. In earlier parallel implementations, the number of processors employed was independent of the size of the scene but depended on the size of the model database (which is usually very large). Our implementations on CM-5 and MP-1 significantly improve upon the number of processors employed and also result in superior time performance. The implementations developed in this paper require a number of processors that are independent of the size of the model database and are scalable with the machine size. Results of concurrent processing of multiple probes are also reported.

UR - http://www.scopus.com/inward/record.url?scp=0011010877&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0011010877&partnerID=8YFLogxK

U2 - 10.1006/jpdc.1994.1044

DO - 10.1006/jpdc.1994.1044

M3 - Article

AN - SCOPUS:0011010877

VL - 21

SP - 96

EP - 109

JO - Journal of Parallel and Distributed Computing

JF - Journal of Parallel and Distributed Computing

SN - 0743-7315

IS - 1

ER -