Ppt

Preview:

Citation preview

Human Mouse Interface

By:ESHPREET BAJWA(Roll no. H-15)

SMITA CHANDURKAR(Roll no. H-27)ANAGHA GOLHAR(Roll no. J-19 )

Under guidance of:Prof. S.A. Khoje

A Project presentation on

INDEX Problem Definition Project Objectives Face detection methods SSR filter Skin color model SVM Face detection algorithm Face tracking H/W & S/W requirements Feasibility Future prospects

Problem definition

This project aims to present an application that is able of replacing the traditional mouse with the human face as a new way to interact with the computer.

Facial features (nose tip and eyes) are detected and tracked in real-time to use their actions as mouse events.

Project Objectives

Real Time Response. To enable physically handicapped people to use the

mouse. Nose tip = mouse pointer. Eye blink = mouse click.

Face detection methods

Two main categories: Feature-based methods, and image-based methods.

Feature-based methods

Image-based methods

Face Detection Algorithm Overview

Face detection general steps.

Find Face Candidates

We will be using feature based face detection methods.

This is to reduce the area in which we are looking for the face, so we can decrease the execution time.

To find face candidates the SSR filter will be used

SSR Filter

SSR Filter stands for: Six Segmented Rectangular filters. The sum of pixels squaren in each sector is denoted as S

along with the sector number.

SSR Filter.

How do we use SSR?

Integral Images

Integral image at location x, y contains the sum of pixels which are above and to the left of the pixel x, y

….contd.

We will need only 3 arithmetic operations to calculate the sum of pixels which belong to it

Sector = D – B – C + A. So each SSR filter requires

6*3 operations to calculate it.

Ideal SSR filter location for a face candidate

Skin Color Model Use the pure r and g values which are the R and G values of the RGB color

model in the absence of brightness, and they are calculated with the following equations:

r = R / (R + G + B) (1) g = G / (R+ G + B) (2) We move the r and g values to the (a, b) color space with the following

equations: a = r + g / 2 (3) b = √3 / 2 g (4) The range of ‘a’ is from 0 to 1, while the range of ‘b’ is from 0 to √3 / 2. 0.49<a<0.59 0.24<b<0.29

SVMSVM stands for: Support Vector Machines, which are new types of maximum margin classifiers.

The center of each cluster that is big enough is set with the following equations:x = [∑ x(i)] / n y = [∑ y(i)] / n i is the pixel from the cluster, n is the cluster’s area.

Find Pupils’ Candidates

Find the pixels that belong to a dark area , the sector is binarized with a certain threshold.

The clusters of the binarized sector are found.

Extract BTE Templates

After finding pupils candidates for each of the clusters (face candidates) we can extract BTE templates in order to pass them to the support vector machine.

We train our support vector machine on templates of size 35 * 21 pixels.

35 * 21 pixels that represent the original template.

An extracted template that was rotated back to a horizontal pupils’ alignment.The larger version is only for observation. It will not be used in our face detection process.

Face Tracking

Detecting The Eyebrows Eyes Tracking Tracking The BTE Tracking the nose tip Blink Detection

Hardware Requirementso Pentium IV processoro Webcam

Software Requirementso Windows XPo JAVA/JAVAX

Feasibility Technical Feasibility

Economical Feasibility

Operational Feasibility

Schedule Feasibility .

Future Works

o Improving the tracking robustness against lighting conditions.

o Adding the double left click and the drag modeo Adding voice commandso Enable/disable controlling the mouse with the face

Thank you