rar

Robustness to changes in illumination conditions as well as viewing perspectives is an important requirement

  • 2014-03-05
  • 89.19KB
  • Points it Requires : 2

Robustness to changes in illumination conditions as well as viewing perspectives is an important requirement form any computer vision applications. One of the key factors in enhancing the robustness of dynamic scene analysis that of accurate and reliable means for shadow detection. Shadow detection is critical for correct object detection in image sequences. Many algorithms have been proposed in the literature that deal with shadows. However, a comparative evaluation of the existing approaches still lacking. In this paper, the full range of problems underlying the shadow detection are identified and discussed. We classify the proposed solutions to this problem using axonomy off our main classes, called deterministic model and non-model based and statistical parametric and non-parametric. Novel quantitative (detection and discrimination accuracy) and qualitative metrics (scene and object independence, exibility to shadow situations and robustness to noise) are proposed to evaluate these Classes of algo-rithms on a benchmark suite of indoor and outdoor videosequences.

unfold

You Might Like

Uploader
hellopinkgirls
 

Recommended ContentMore

Popular Components

Just Take a LookMore

EEWorld
subscription
account

EEWorld
service
account

Automotive
development
circle

About Us Customer Service Contact Information Datasheet Sitemap LatestNews


Room 1530, 15th Floor, Building B, No.18 Zhongguancun Street, Haidian District, Beijing, Postal Code: 100190 China Telephone: 008610 8235 0740

Copyright © 2005-2024 EEWORLD.com.cn, Inc. All rights reserved 京ICP证060456号 京ICP备10001474号-1 电信业务审批[2006]字第258号函 京公网安备 11010802033920号
×