Portal de Periódicos da CAPES

An analysis of out-of-plane facial detection using variant dataset among UTEM students by using speed-up robust feature technique

2016; American Institute of Physics; Volume: 1774; Linguagem: Inglês

10.1063/1.4965099

1935-0465

Mohd Safirin Karis, Nursabillilah Mohd Ali, Nur Warrissyahidah Badrul Hisham, Amar Faiz Zainal Abidin, Nurul Fatiha Johan, Nur Rafiqah Abdul Razif,

Digital Media Forensic Detection

The core objective of this project analysis is to determine out-of-plane facial detection by using variant dataset of SURF technique from the captured image of different angles. Mainly purpose of this technique is to ascertain the number of crimes that can be recorded by CCTV where detecting face image is problematical. Since face is the most crucial aspect for identification, many images captured are out-of-plane as compared to in-plane images respectively. Therefore, this system would be able to identify and detect face image according to its spectrum, range and scale. Hence, the algorithm method will be applied by using Microsoft Visual Studio 2013. This project analysis is divided into two stages; the first stage of part 1: is to study about basic SURF technique, the second stage of part 1: is to analyse out-of-plane facial detection. The sample of dataset is collected among females and males Malay students of age between 20-25 years old. Five different angles of each sample is analysed comprehensively in order to acquire exact results by using true positive image based on the key features detected. Hence, an image as well as the process time is a focal aspect as well as the extraction time in one image also is a property to be developed and upgraded.