Menu
Home
Log in / Register
 
Home arrow Computer Science arrow Virtual, Augmented and Mixed Reality
< Prev   CONTENTS   Next >

2.2 Visual Sensitivity of Human Perception

A conventional approach to evaluate the attraction of objects in an image is based on textural information. In this direction, regional structural analysis algorithms based on gradient are used to detect features. However, saliency is considered better to reflect sensitivity of human vision to certain areas on an image thus benefits context awareness systems [6]. Visual saliency [7], human perceptual quality indicating the prominence of an object, person, or pixel to its neighbors thus capture our attention, is investigated by multiple disciplines including cognitive psychology, neurobiology, and computer vision. Salient maps are topographical maps of the visually salient parts of scenes without prior knowledge of their contents and thus remains an important step in many computer vision tasks.

Saliency measures are factors attracting eye movements and attention such as color, brightness, and sharpness, etc. [8]. Self-saliency is a feature that expresses the inner region complexity, which includes color saturation, brightness, texture, edginess, etc. Whereas, relative saliency indicates differences bet1een a region and its surrounding regions such as color contrast, sharpness, location, etc. Saliency measures can be combined with different weights to determine important regions more efficiently.

Most of saliency object detection techniques can be characterized as bottom-up saliency analysis, which is data-driven [9], or top-down approach, which is task-driven [6]. We focus on pre-attentive bottom-up saliency detection techniques. These methods are extensions of expert-driven human saliency that tends to use cognitive psychological knowledge of the human visual system and to find image patches on edges and junctions as salient using local contrast or global unique frequencies. Local contrast methods are based on investigating rarity of an image region with respect to local neighborhoods [8]. Whereas, global contrast based methods evaluate saliency of an image region using its contrast with respect to the entire image [10].

In this paper, the authors propose an efficient based human vision computation method to detect automatically high informative regions based on regional contrast in order to determine which region contains meaningful keypoint candidates. This reduces redundant candidates for further processing steps.

 
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Philosophy
Political science
Psychology
Religion
Sociology
Travel