Knowledge Discovery by Image Recognition in Robots

8/7/2019 Knowledge Discovery by Image Recognition in Robots

http://slidepdf.com/reader/full/knowledge-discovery-by-image-recognition-in-robots 1/7

A Technical paper presentation on the topic

KNOWLEDGE DISCOVERY BY IMAGE

RECOGNITION IN ROBOTS Authorised By

SANTOSH BHARADWAJ REDDY

Email: [email protected]

Engineeringpapers.blogspot.com

More Papers and Presentations available on above

site

Abstract:

Till now we have the ROBOTS

which will only execute the program and run

those parts according to that. But, my

project aim is to provide the social

intelligence, common sense to the robots i.e.

we have to design the ROBOTS that will

make decision itself according to the

situation based on how the things are going

on surrounding it. The robot that acts similar

to the human brain. For example, If there is

no one in the room in which AC is ON, the

robot must be able to switch OFF the AC to

save the power, and many more as human

brain takes decision. In this paper we will

discuss about the small topic in my whole

project i.e. human face recognition

technique.

Introduction:

Definition of a Robot

• "A reprogrammable, multifunctional

manipulator designed to move material,

parts, tools, or specialized devices through

mailto:[email protected]




various programmed motions for the

performance of a variety of tasks" .

• Or a simpler version

• An automatic device that performs

functions normally ascribed to humans or a

machine in the form of a human.

What is a Robot

• The term robot derives from the Czech

word robota, meaning forced work or

compulsory service, or robotnik, meaning

serf.

Sensors:

• Sensors that tell the robot

position/change of joints: odometers,

speedometers, etc.

• Force sensing. Enables compliant

motion--robot just maintains contact

with object (video: compliant)

• Sonar. Send out sound waves and

measure how long it takes for it to be

reflected back. Good for obstacle

avoidance.

• Vision systems

Effectors:

• Conve

rts

software commands into physical

motion

• Typically electrical motors or

hydraulic/pneumatic cylinders

• Two main types of effectors:

locomotion

manipulation

Locomotion:

Legs!

traditional (video: honda human)

Other types Statically stable locomotion:

can pause at any stage during

its gate without falling

Dynamically stable

locomotion: stable only as

long as it keeps moving

(video: hopper)

• Still, wheeled or tread locomotion

like Shakey is still most practical for

typical environments

• Other methods: reconfigurable

robots, fish robots, snake-like robots.

(video: mod-robot)

Manipulation:

Manipulation of objects

Typical manipulators allow for:



Prismatic motion (linear

movement)

Rotary motion (around a

fixed hub)

• Robot hands go from complex

anthromorphic models to simpler ones

that are just graspers

(video: manipulation)

(video: heart surgery)

There are two basic problems for the robots

to work according to the environment. They

are:

1) Localization and Mapping

2) Motion Planning

Localization:

Finding the location where it is

right now????

How Do We Solve Localization??

There are many algorithms to

solve Localization problems. They are:

1) Markov localization (simplest)

2) Kalman filters (historically most

popular)3) Monte Carlo localization /

particle filters

Mapping:

Finding map of the environment

of the robot by using implicit sensors

(location of the robot as it moves).

Motion Planning:

• Simplest task that a robot needs to

accomplish

• Two aspects:

Finding a path robot should

follow

Adjusting motors to follow

that path

• Goal: move robot from one

configuration to another.

AssumptionThis part of the paper deals with the

basic idea of us i.e. recognizing the human

faces and providing the information about

the person. This is done by the process as

given below:

First, if a person stands before the

robot and talked with that then while talking

to the human, robot follows a procedure of

retrieving the required data. The procedure

is:

1) The robot activates the recording

device i.e. camera to record the



video of the human being behind

to it.

2) It recognizes the human face by

using one of the face recognition

techniques like video based

tracking and recognition.

3) After identifying the human face

it checks the database for the

same face that is previously

stored.

4) If the image is already exists then

it recollects the information

about the person that is stored

previously in the last interaction.

5) If it is not present then it follows

the following steps:

a.First it asks for the name of the

person.

b. Stores the face with the name

of that person in the database.

c.Creates the instance to the

variable which can store a large

amount of the data related to

that person.

d. It formally listens to the

person and converts the voice

into text.

For all the above discussion consider an

example of the conversation of the human

being with the robot:

In the above figure, a person is talking to the

robot: when he said Hi! to the robot then in

response the robot will also say Hi! with

some expression and body movements

accordingly as he moves. Mean time, it

records the face of that person and checks

the database for the exact face, if it is not

present in the database then it stores the face

of that human being by asking the name of

that person as:

“May I know your Name Please?”

Then in response the person will tell his

name as: “Eigen ville” then it store the face

with the name of that person and create an

instance to store further information

regarding to that person.

And further the interaction may be done

like this:



Man: “what’s your name?”

Robot: “this is Android, from where you

are?”

Man: ……..

……….

With this conversation according to the

keywords used by the human being it will

stores the important information like Name,

Place, Age, Education, and Personal Details

and even according to the conversation and

reaction of the person it can also come to

estimation about the character of that person

and accordingly him it will give the

response to the interaction.

This is what we mean as

commonsense, as human brain acts

according to the situation. Similarly, we can

design the robot as providing the ability to

act according to the situation.

Here the face recognition is done

from a video using the Hue value. The

process is according to this:



In this paper we used only the face

recognition technique. And we have many

more topics to cover to run the robot

according to our aim. The concepts that we

have to cover are: the face comparison

techniques, technique used in Voice to Text

converter, techniques used to provide the

commonsense, video database i.e. storing

the pictures and videos for further recall of

the situation and many more.

Conclusion:

In this paper we have covered the

topics like important points about the robots,

use of robots, purpose of robots main parts

of robots, problems facing for automation of

robots, technique used to recognize the

human face. With this paper we came to an

idea on our aim to provide the information

about the persons that are talking to robot.

References:

• Automatic Face Recognition Using

Color Based Segmentation and

Intelligent Energy Detection,

Michael Padilla and Zihong Fan.

• Robotics, Dr.Julie and A. Adams

from Vanderbilt University.



Authorised By

SANTOSH BHARADWAJ REDDY

Email: [email protected]

Engineeringpapers.blogspot.com

More Papers and Presentations available on above

site



Date post:	09-Apr-2018
Category:	Documents
Upload:	bharadwaj-santhosh
View:	217 times
Download:	0 times

Knowledge Discovery by Image Recognition in Robots

Documents