Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

How to detect shape and size of an object without camera?

Status
Not open for further replies.

xpress_embedo

Advanced Member level 4
Joined
Jul 5, 2011
Messages
1,154
Helped
161
Reputation
396
Reaction score
189
Trophy points
1,353
Location
India
Activity points
10,591
Hello! Everyone,

I want to detect the shape and size of the object placed inside a box of 30x30x30cm.
Is there any sensor or sensors arrangement with which i can sense these things and plot the graph on PC.

I searched and found that Ultrasonic Sensor can be used to sense distance and by placing sensor all over and using stepper motors to sense the complete boundary of the object, and then plotting the values in MATLAB or any other Software.
Is this is fine?
Please suggest any other good method, which can be used to create a replica or the object on PC, if i get the coordinates of the object boundary, i will reconstruct the image on PC.

Thanks in advance.
 

As far as I know, it is possible to build 3D scanners with laser arrangements, and has a good resolution to take details form the target object.
 
Project an image through a convex lens onto a grid of photosensors, say 10x10. It's like making your own digital camera. Resolution is crude but it should tell you whether an object is round or square or triangular. By adjusting range of response you might even get shades of gray. Build a second assembly if you want stereo vision, to get more clues as to shape and size.

Obviously you'd like to get better resolution, although it requires more photosensors. It can become hard to work with. There may be suitable manufactured devices which are easier to work with than the CCD in today's digital camera.
 
As far as I know, it is possible to build 3D scanners with laser arrangements, and has a good resolution to take details form the target object.

Thanks for the information, i searched with this on internet, and saw couple of videos and it looks nice to me, the only problem in this method, is that we have to rotate the object, which is not possible in my case as the object is stationary, and i think instead of that i can rotate the laser sensor to complete the 360 degree scan.

Is 360 degree rotation of sensor and receiver is possible?
Can you please provide me the part numbers of few laser sensor and receiver?

- - - Updated - - -

Project an image through a convex lens onto a grid of photosensors, say 10x10. It's like making your own digital camera. Resolution is crude but it should tell you whether an object is round or square or triangular. By adjusting range of response you might even get shades of gray. Build a second assembly if you want stereo vision, to get more clues as to shape and size.

Obviously you'd like to get better resolution, although it requires more photosensors. It can become hard to work with. There may be suitable manufactured devices which are easier to work with than the CCD in today's digital camera.

The box containing the object doesn't have light, so photo transistors will not work, and to make them work i have to provide the light source from the opposite side, well that is possible and i can give it a try.
Can you please tell me what is role of the Convex Lens in this application?

I have to read about CCD, and from the post it looks that, either i have to use CCD or Photo-Transistor/Diode.
 

Can you please tell me what is role of the Convex Lens in this application?

AKA magnifying glass. The convex lens is the part that makes a camera possible. It creates a focussed image, upside-down.

A dimestore plastic reading glass probably will do the job. A couple inches diameter, a couple inches focal length. That should project a usable image.
 
I would suggest that you stick to using laser sensors, there is some types of laser sources that can send out multiple beams simultaneously in different directions to cover the whole object and you can rely on the measured material reflectivity at most times as where reflected beam differs it would be the border. i don't recall how these work but i think it's a good solution.
 
I believe they spin a prism- see supermarket checkouts.
 

I would suggest that you stick to using laser sensors, there is some types of laser sources that can send out multiple beams simultaneously in different directions to cover the whole object and you can rely on the measured material reflectivity at most times as where reflected beam differs it would be the border. i don't recall how these work but i think it's a good solution.

Can you please share the part numbers of such sensors including transmitters and receivers.
And its approximate cost.
 


There is a compromise between the quality of the image to be acquired and the complexity of the solution. Unless you have previously notion of the shape of the object (say, a set of pre-known shapes) there is no other way of obtaining stereoscopic vision without or rotating the object, or having more than one sensor / reflector. In my opinion, the solution based on camera and / or vision system is the most suitable for that purpose.
 
I wonder how expensive terahertz technology is these days...
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top