Performance Evaluation of a Vision System for Automatic Object Recognition

QSpace/Manakin Repository

Performance Evaluation of a Vision System for Automatic Object Recognition

Show simple item record


dc.contributor.author Bennamoun, M
dc.contributor.author Boashash, B
dc.date.accessioned 2012-02-15T05:53:37Z
dc.date.available 2012-02-15T05:53:37Z
dc.date.issued 1998
dc.identifier.citation Applied Signal Processing (April 1998), 5 (2), pg. 62-81 en_US
dc.identifier.uri http://hdl.handle.net/10576/10782
dc.description This paper presents a complete analysis of a new 2D vision system previously developed with performance evaluation. (Additional details can be found in the comprehensive book on Time-Frequency Signal Analysis and Processing (see http://www.elsevier.com/locate/isbn/0080443354). In addition, the most recent upgrade of the original software package that calculates Time-Frequency Distributions and Instantaneous Frequency estimators can be downloaded from the web site: www.time-frequency.net. This was the first software developed in the field, and it was first released publicly in 1987 at the 1st ISSPA conference held in Brisbane, Australia, and then continuously updated). en_US
dc.description.abstract This paper presents a comprehensive analysis of a vision system algorithm previously developed by the authors and examines its performance in terms of accuracy, robustness, and efficiency. The system accuracy and robustness are evaluated under the presence of variations in conditions such as noise, partial occlusion, scaling, translation and rotation. Efficiency of the system is measured in terms of the speed and the amount of memory required. The system was extensively tested on a simulation program built on the Khoros image processing software. The results showed that the suggested vision system is invariant to translation, rotation, and scaling and robust to noise and partial occlusion. An accurate detection is achievable at 3 dB signal-to-noise ratio and when 25% of the object boundary is occluded. The system is highly efficient due to the small memory requirement as well as the fast and simple matching process. Moreover, the use of the modular approach in the overall algorithm increases the system flexibility. The algorithm of each block can be expanded or replaced with a better one if such an improvement is required. en_US
dc.language.iso en en_US
dc.publisher Springer-Verlag London en_US
dc.subject Automatic object recognition en_US
dc.subject Convex dominant points en_US
dc.subject Convex point en_US
dc.subject Invariance en_US
dc.subject Part segmentation en_US
dc.subject Super-quadrics en_US
dc.subject Structural description en_US
dc.subject Vision systems en_US
dc.title Performance Evaluation of a Vision System for Automatic Object Recognition en_US
dc.type Article en_US

Files in this item

Files Size Format View
Boashash-Bennam ... ing_Performance-Vision.pdf 5.689Mb PDF View/Open

This item appears in the following Collection(s)

Show simple item record

Search QSpace


Advanced Search

Browse

My Account