Face anonymiser

11 Jan 2014
Face anonymiser for images and videos
Total votes: 2
Face anonymiser for images and videos

Face anonymiser for images and videos - Umar Spa

In this work we have implemented face anonymisation in images and videos. Initially we have tried to explain Viola Jones algorithm for face detection, focusing particularly on three specific innovative features of this algorithm, i.e. "Haar-like features", "Integral image" and "Adaboost". It’s the combination of these features that make this algorithm extremely rapid and robust. Then we have discussed the implementation details of Viola - Jones algorithm along with the testing of the proposed solution on two different databases, in order to determine the best parameters of the function used in the implementation process and to evaluate the efficiency of the proposed solution. And then, finally we have implemented the face anonymisation on images and videos by applying a blurring effect on the detected faces using two different techniques: a default function taken from a library at our disposal and a new function proposed by us. Both the techniques have been implemented on images, as well as on videos.

 

 

Thesis: 
References: 

<pre style="color: rgb(0, 0, 0); word-wrap: break-word; white-space: pre-wrap;">
<span style="font-family: Arial, sans-serif; font-size: 13px; line-height: 16px;">Viola, Paul, and Michael J. Jones. &quot;Robust real-time face detection.&quot;&nbsp;</span><i style="font-family: Arial, sans-serif; font-size: 13px; line-height: 16px;">International journal of computer vision</i><span style="font-family: Arial, sans-serif; font-size: 13px; line-height: 16px;">&nbsp;57.2 (2004): 137-154.</span></pre>