Deidentifikacija lica preinačenim postupkom k-istih lica upotrebom karakterističnih točaka lica

master's thesis

• Title (english): De-identification of Faces by Using Modified k-Same Method and Characteristic Facial Points
• Author: Matej Crnac
• Mentor: Slobodan Ribarić (mentor)
• Committee member: Slobodan Ribarić (predsjednik povjerenstva)
• Committee member: Marin Golub (član povjerenstva)
• Committee member: Vlado Sruk (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2018-07-16, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad bavi se problemom deidentifikacije lica. Razvijen je program koji radi deidentifikaciju preinačenim postupkom k-istih lica te koristi karakteristične točke lica. Predstavljene su tri metode za računanje udaljenosti između dva lica, a metoda temeljena na određenim udaljenostima se pokazala kao najbolja. Uz to je napravljena i klasifikacija spola kako bi deidentifikacijom osoba ostala istog spola. Razvijena metoda deidentifikacije računa udaljenost lica od svih lica iz galerije slika, te zatim, ovisno o parametru k, radi zamjenu dijelova lica s tom ili više slika. Na kraju su se rezultati ispitali testiranjem na ljudima koje je pokazalo dobre rezultate kao i moguća poboljšanja.

Abstract (english)

The topic of this work is face de-identification. A program was developed to do de-identification by using a modified k-Same method and characteristic facial points. Three methods for calculating distance between faces were developed and the method based on specific distances was proven to be the best. Gender detection method was also developed because it can be important for de-identified person to stay the same gender. De-identification method is calculating face distance from every other face in the gallery and then, based on parameter k, does swapping face parts with one or more faces. In the end, results were tested on people and results are proven to be good with some possible improvements.

• Keywords: deidentifikacija lica; detekcija lica; detekcija karakterističnih točaka lica; udaljenost lica; afina transformacija; klasifikacija spola; zamjena dijelova lica
• Keywords (english): face de-identification; face detection; facial landmarks detection; distance between faces; affine transformation; gender detection; swapping facial parts
• Language: croatian
• URN:NBN: urn:nbn:hr:168:151013
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-03-13 19:36:53