Selection of canonical images of travel attractions using image clustering and aesthetics analysis
by Jen-Chang Liu; Yin-Chen Liang; Shih-Wei Lin
International Journal of Computational Science and Engineering (IJCSE), Vol. 8, No. 4, 2013

Abstract: The popularity of smart phones and development of 3G mobile networks have dramatically increased the sharing of photos on social platforms. However, the huge numbers of images depicting tourist attractions uploaded to photo-sharing websites vary in terms of subjective photographic intent and contain considerable environmental noise. We propose an approach for finding canonical images of travel attractions from online social platforms, and apply aesthetics analysis to rank the results. This approach can provide travel websites with a mechanism for automatic image selection and help travellers browse travel spots. The methods used include face detection to filter attractions images containing people, feature extraction, and feature classification to filter out background features. We then calculate the similarity among images, and apply an affinity propagation algorithm for clustering and find canonical images. Finally, the clustered images are ranked by aesthetics scores. Experimental results show that the proposed approach obtains representative and aesthetically-pleasing images for attractions that include artificial landmarks.

Online publication date: Fri, 27-Dec-2013

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Computational Science and Engineering (IJCSE):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com