Abstract
Object-based visual attention has got more and more attention in image processing. A computational model of visual attention based on space and object is proposed in this study. Firstly spatial visual saliency of each pixel is calculated and edges of the input image are extracted. Salient edges are obtained according to the visual saliency of each edge. Secondly, a graph-based clustering process is done to get the homogeneity regions of the image. Then the most salient homogeneity regions are extracted based on their spatial visual saliency. Perceptual objects can be extracted by combining salient edges and salient regions. Attention value of each perceptual object is computed according to the area and saliency. Focus of attention is shifted among these perceptual objects in terms of the attention value. The proposed computational model was tested on lots of natural images. Experiment results indicate that our model is valid and effective.
Highlights
Human and other primates can find the objects of interest in complex visual scenes quickly
Attention value of each perceptual object is computed according to its average visual saliency and its location and size
To evaluate the performance of the proposed computation model of selective visual attention, we have tested it in many natural images
Summary
Human and other primates can find the objects of interest in complex visual scenes quickly. Most of the current methods compute the attention value only based on the average saliency of the pixels in the perceptual objects. Attention value of each perceptual object is computed according to its average visual saliency and its location and size. When the number of salient pixels is larger than 70% of the area of the whole image, the weight of the feature saliency map is set to zero. This means it is not included when in feature integration. Calculate saliency value of each pixel based on feature value contrast and generate saliency map of the input image. The example of perceptual object extraction and attention shift will be shown in below section
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: Research Journal of Applied Sciences, Engineering and Technology
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
