Badly need some help here!

[hyma]

The aim of the project is to design a system which would automatically focus onto a circular black cardboard(of about 10 cm diameter) placed on a 2m x 2m white wall and then automatically fire a laser to hit the target(i.e., the black cardboard). The device is to be placed at a distance of 10 m from the white wall.

I came up with a rough plot of it which I've jotted down below.

1. Take the photo of the entire 2mx2m white wall located at a distance of 1m from the device.

2. From the image, locate the black target,provided within the white background wall, and obtain it's spatial orientation(it's 3d coordinates).

3. Find a mechanism to orient the system and thus align it in accordance with the position vector so obtained in step 2 and thus focus on the target.

4. The system must then automatically fire a beam of laser to hit the target.

This was what I could best think of.
Now I wonder if there are any relevant techniques to achieve them.

I shall be pleased to know if there are any simpler ways to solve this.

Thanks in advance!

 

[Mr RB]

What's it for?

An open loop system will *barely* be accurate enough if the camera, laser and wall are always permanently fixed in place.

It would be better to use a closed loop system where the laser is modulated at low power, then the camera can "see" both the laser spot and the target spot at the same time. Then use a closed loop system to bring the laser onto target.

 

[hyma]

The target can be placed at any random position on the white wall. The laser is to be mounted on the device which would align itself inline with the target(the black cardboard).

 

[Boncuk]

Use a stereographic and autofocussing camera to calculate angles.

 

[hyma]

Search for "stereograph" returned me to stereoscopy page. Couldn't understand how it can measure angles.

 

[hyma]

What's it for?

An open loop system will *barely* be accurate enough if the camera, laser and wall are always permanently fixed in place.

It would be better to use a closed loop system where the laser is modulated at low power, then the camera can "see" both the laser spot and the target spot at the same time. Then use a closed loop system to bring the laser onto target.

"The camera can "see" both the laser spot and the target spot at the same time."Can you please elaborate a little on that?

 

[ericgibbs]

hi,
A low cost PC webcam, with IR filter removed.
Tap into the hori/line sync signals and video signal, look for the minimum video signal that occurs as the 'black target' is scanned.
This will give you an indication of the X,Y coord on 2*2 mtr plane.
Use these coords in a PIC controller to drive the aiming laser pointer.
or
Using the same PC webcam with an aperture iris that allows a very small angle of view, scan the 2*2m plane from top left to bottom right as a X,Y raster, the same mechanism drives the aiming laser
As the video signal falls to near zero on the black target, fire the aiming laser.

What you are trying to do is not easy.

 

[hyma]

hi,
A low cost PC webcam, with IR filter removed.
Tap into the hori/line sync signals and video signal, look for the minimum video signal that occurs as the 'black target' is scanned.
This will give you an indication of the X,Y coord on 2*2 mtr plane.
Use these coords in a PIC controller to drive the aiming laser pointer.
or
Using the same PC webcam with an aperture iris that allows a very small angle of view, scan the 2*2m plane from top left to bottom right as a X,Y raster, the same mechanism drives the aiming laser
As the video signal falls to near zero on the black target, fire the aiming laser.

What you are trying to do is not easy.

"Tap into the hori/line sync signals and video signal, look for the minimum video signal that occurs as the 'black target' is scanned."
I couldn't get this. What are the hori/line sync signals all about?

 

[ericgibbs]

"Tap into the hori/line sync signals and video signal, look for the minimum video signal that occurs as the 'black target' is scanned."
I couldn't get this. What are the hori/line sync signals all about?

hi,
Read up on TV raster scanning.
Raster scan - Wikipedia, the free encyclopedia

Is this going to a practical project or just a theoretical project report.?

 

[hyma]

A practical one...
And the laser system must be able to automatically turn itself on as soon as it locates the target. Then I thought if the laser system is mounted on some rotating base, whose angular rotation is determined by the position of the black spot, it can then be made to hit the target. I wonder if a stepper would serve the purpose.

In your post, u suggested "the same mechanism drives the aiming laser". Is there anyway that I fire the beam,without using any stepper or the like, right onto the target?

 

[ericgibbs]

A practical one...
And the laser system must be able to automatically turn itself on as soon as it locates the target. Then I thought if the laser system is mounted on some rotating base, whose angular rotation is determined by the position of the black spot, it can then be made to hit the target. I wonder if a stepper would serve the purpose.

In your post, u suggested "the same mechanism drives the aiming laser". Is there anyway that I fire the beam,without using any stepper or the like, right onto the target?

hi,
You must have some means of aiming the laser onto the target.

If you use the same steppers/etc to turn the scanning camera and the aiming laser, then the aiming laser will be pointing at the target when its detected.

 

[hyma]

That means even the camera has to be mounted onto the same stepper. But then, will it be possible to define the boundaries of the area that has to be scanned?

 

[ericgibbs]

That means even the camera has to be mounted onto the same stepper. But then, will it be possible to define the boundaries of the area that has to be scanned?

The PIC or MCU will drive the X,Y steppers within the coordinates of the 2m^2 white background located 10mtrs away, its just basic trig.

EDIT:
By using a collimator over the camera lens, so that the camera can only see a 20mm dia area, you will not need to use the camera raster signals, just the video signal

 

[hyma]

A correction to my post there!

The target is 1m* (and not 10 m ) away from the wall.

 

[ericgibbs]

A correction to my post there!

The target is 1m* (and not 10 m ) away from the wall.

In that case will the WHITE background of the BLACK target be well lit.?

If yes, you could just a photo detector/amp with a simple lens and collimator in place of the camera.

When the output from the PD/amp falls close to zero, fire the laser.

 

[hyma]

The white background may be a plastic board or just a plane wall. And it is also given that a webcam and a few image processing algorithms are to be used for this purpose.

 

[ericgibbs]

The white background may be a plastic board or just a plane wall. And it is also given that a webcam and a few image processing algorithms are to be used for this purpose.

OK,
At least we agree on the webcam method, whats your next step.?

 

[hyma]

On locating the target, there should be some means by which the laser system is turned on. This would probably be possible by using some relevant code that sends appropriate control signals to the laser system which
i. Turn the system on.
ii. Orient the system inline with the target.
iii. Fire the laser beam on locating the target.

But then, will it be accurate enough? I'm not sure of the practical constraints that have to be faced. Planning to visit the nearest dealer in electrical parts this weekend.

I find things simple but yet complicated..

Thanks for your support, Sir!

 

[ericgibbs]

On locating the target, there should be some means by which the laser system is turned on. This would probably be possible by using some relevant code that sends appropriate control signals to the laser system which
i. Turn the system on.
ii. Orient the system inline with the target.
iii. Fire the laser beam on locating the target.

But then, will it be accurate enough? I'm not sure of the practical constraints that have to be faced. Planning to visit the nearest dealer in electrical parts this weekend.

I find things simple but yet complicated..

Thanks for your support, Sir!

I guess the requirement is to hit somewhere on the target, not dead centre.

you could feed the video signal [filtered] to an analog input on the PIC, use that to say fire laser.

 

[hyma]

Oops! Missed a part there..

In the Webcam method that you've suggested to scan for the target, will time be a constraint?
And one of my friends suggested to take a JPEG file of the image and scan for the code that corresponds to black. Is any such thing possible?