Automatic profile picture validation

Matic Vrščaj (2017) Automatic profile picture validation. EngD thesis.

Preview

Abstract

The purpose of this study was to explore areas of computer vision, which may prove useful in automatic profile picture validation in Information Systems. For practical example, we used a web portal that enables users to upload their own profile pictures. In the first part we explain useful algorithms for face detection, age and gender. The second part focuses on implementation of our own solution in Java, which uses explored algorithms to suggest the appropriateness of the image based on the detected facial features - age and gender. An image can be marked as appropriate, if it contains exactly one face of an adult and is not blurred. This work overlooks image quality but focuses on face detection. Finally, we calculate success rate, the likelihood that the algorithm can determine the suitability of random user image, and discuss how our approach could be improved.

Item Type:

Thesis (EngD thesis)

Keywords:

computer vision, face detection, gender detection, age detection

Number of Pages:

Language of Content:

Slovenian

Mentor / Comentors:

Name and Surname	ID	Function
viš. pred. dr. Borut Batagelj	298	Mentor

Link to COBISS:

http://www.cobiss.si/scripts/cobiss?command=search&base=51012&select=(ID=1537379523)

Institution:

University of Ljubljana

Department:

Faculty of Computer and Information Science

Item ID:

3814

Date Deposited:

03 Mar 2017 09:32

Last Modified:

14 Mar 2017 11:37

URI:

http://eprints.fri.uni-lj.si/id/eprint/3814

Actions (login required)

View Item