Doc

ranchocucamongabrrrAI and Robotics

Nov 6, 2013 (4 years and 1 month ago)

116 views


1

Module
7
:

Face

Recognition

In this
module, you will learn:



How to training a face image.



How to make Robot recognize the man whose face image is saved.



Requirement:



Knowledge about


computer vision



OpenCV library documents



Eclipse is installed Android
SDK, ADT.



Copy robot sdk before using Eclipse.



Content:



Task 1: Training face image.



Task 2: Recognize a face.















2

Face recognition module on the NAO robot can recognize the face that has been remembered
before.


Module consists of two processes
are

training data and face
recognition

Task:
Training face image

This process can be independent of the robot. You can send
some

face

p
icture
s

which was

taken
in good lighting and clean
ing background

up
to s
erver
.

S
erver
will use

image processing
algorithm
s to extract and save the face features

in the database then
add in

the registered account
.









User API:



Training
Apis:


Training Face Image



3



Task:
Recognize a face




Detect Api:




















Send data to NAO
-

Use API:

DatabaseTransfer.pushDatabase()


Capture
Image from
NAO’s
牯扯琠
ca浥ma

&

䑥瑥t琠Face


Face Crop


Receive
Face Image

Send Face
Image to
Server


Use Haar like
features
-

OpenCV

Process data
and send
Image
infomation

NAO

Android

Server

Start Module
-

Use API:

RobotFaceDetection
.
startDetection
()


Receive
Image
information


TTS


Motion…

Send data to server Using the above
training Apis



4
























FaceDetection processing use
OpenCV



Haar like features



Crop one Face Image

Face

requirements



Vertical



Good light

Send Face Image to
cloud or
server

Receive Image
information



Text to speech



Behavior …

Send Face Crop
Image to android

Start, Stop

Send request…