Image processing in Robot

[$@$]

Hi...can anybody help me in my final project...
I want to build a tracking robot with image processing censor...
I mean this robot can mapping the area...
First the robot explore the area...it can sense the wall or barrier with image censor not with IR or ultrasonics... then after it finished, the data can be render to 3D in computer....
   

[sweet back]

what components and accessories do you have access to?

[$@$]

I want 2 use a web-cam...so the robot can avoid the barrier or wall.

Block Diagram:
Web Cam-->Microcontroller-->Servo Motor

I'am really confused about interfacing....from web-cam to microcontroller.....
can anybody help me......

[Ilias]

to put it simply, you just can't have image processing (in robotics is called "machine vision") by using a cmos sensor (that's the camera) and microcontroller. there are a lot of tasks involved.
to get you started check this link out
http://emsys.denayer.wenk.be/?project=emcam&page=cases&id=16

For image processing you need significant processing power. Some tasks required are compression, frame rate, edge detection, motion detection etc.
There is the "easy" but expensive way and the "hard" and cheap way.
The "easy" way involves the integration of integrated machine vision systems like those in the attachement.
The "hard" way involves a camera that communicates prefferably wirelessly with a pc, which with your own crafted algorithms image processing is performed and the required data are modified accordingly and sent to the robot in order to perform the required action.
In my opinion this is the best way to do it, since you have "unlimited" processing power and several algorithms have already been developed. Additionally it is the cheapest way to do it.
Now if you have other directions by your supervisor we can talk about it again.
It would be great though if you could be more specific about the components you have access to... Do you have to choose your own components, and if yes what is the budget. If the components are already defined, what are they?
Remember that earlier mobile robots with computers on board needed almost 15 minutes to calculate the next step. So response time is another significant factor. Consequently for fast response times off-board processing is compelled.

[Ilias]

to conclude my suggestion is
webcam-->pc-->mcu-->actuators

[Ante]

I don’t know if this will help but the “Where am I sensor and methods for mobile robot positioning” will be posted in Resources!

[Shahriar]

Dear $@$

sometime ago I decided to build a line follower robot by a small camera.( Not a Web cam.)
I had decided to detect between the black Line and the white background by proceesind the video sinal. and to position it by other pulses in a component video signal.

U should have enough information about a video signal to understand what I mean.

Although i didn't the job, But I still think it is possible to do such thing very easily. It has many benefits to using a sensor. meybe u can use suck tricks to recognize wall  depending on your conditions.

HTH - Shahriar

[$@$]

Thx a lot..
But i'm still confused about image processing in it....
What algoritm should i use to avoid the barrier (obstacle avoidence)?
Bcoz the camera didn't know if the barrier in front if it or not? (only 2D)......
so the next quesion--> how we can measure the distance between camera and the barrier?(coz the data only in picture or video)............i'm just using camera sensor......

[Ilias]

do you have any hints from your supervisor???

A good place to start looking is the seattle robotics society.
http://www.seattlerobotics.org

In general
1. you get the image
2. you extract features of the image that you are interested in, most usually edges which are apparent because of the colour difference of the object and the background color.
3. On the next move of the robot you compare the current image with the previous one and if the image is larger that the one before (that is if the same edges are more spreaded on the new image) then you are probably closer to that object.

Take a look at the attachement and don't forget to look more forums as the seattle robotics society one...What are the project specifications??? for example are you supposed to used an 8-bit microcontroller or an arm-based one?

http://www.seattlerobotics.com/cmucam.htm
http://www.cis.udel.edu/~cer/arv/
google "robot vision"


now in order to find source code paradigms and how others have done it google for "source code". it will give you results such as www.koders.com and you can browse through ready made projects or other source codes on relevant projects

[$@$]

What the best method :
1. Single Camera (Monocular)
2. Double Camera (Stereo)
to measure distance between mobile robot and obstacles?? (can it possible with camera???)

should i capture / sampling the picture(jpeg or bmp) from the video (fps-->frame/second).....i mean the filtering process should in pictures or video....

[Ilias]

http://en.wikipedia.org/wiki/Computer_vision

[Ilias]

have you ever used Linux... Well I've got an application from a book of mine, that it is suppose to compile for Linux. I don't think that windows will be a problem....try Borland's C compiler or any other compiler you prefer.
Please do mind that any use of the code uploaded here should be fully referenced.
The web page included (0750677783.html) is an amazon link to the book...Go to the web site of the author. you will probably find more info.

you will probably want to look here->http://www.aaai.org/AITopics/html/vision.html
and also try "googling" for "open source robotics"

[Ilias]

$@$....
How is is going...
Do you need any help compiling the code?
Tell me about the progress of the project....

[$@$]

I'm getting stuck for a month....coz my lecturer told me to add the filter...
i mean that the image must be 'normalized' first before it can be processed..

Image-->Filter-->PC

eg. :The light in the indooor room is 5 lamp, so if the camera capture the image in the room that has 10 lamp-->image must be normalized based on the room that has 5 lamp....(reduced brightness and contrast)

I already found the filter to reduced it, like histogram equalization; edge detection; but how to implemented it in real time image processing??