Dear All,
I am looking for a visual system to connect it to a PLC . Anyone with experience on using visual system has any recommendations ?

Thank You.

something like this ( for automation ):

 

"visual system" is extremely vague, no meaningful answer can be given without more information.

 

Double post?
Max.

 

Double post?
Max.

nope not for me .

 

Check out the products from SICK... great stuff.. cool software
http://www.sick.com/us/en-us/home/products/product_portfolio/vision/Pages/SICK_Vision.aspx

You didn't give any details though on what exactly you are intending to do..

 

There is OpenCV? But I don't know how to connect it to an PLC. You may need another microprocessor to run it.

 

I just did a vision system with OpenCV. Very powerful system, very easy to use for a good programmer and someone experienced with image analysis and segmentation. But djsfantasi is correct, you'd have to customize an interface between your target computer and the PLC.

If you're committed to an industrial PLC, something like the Cognex or SICK would be the way to go. I've had a lot of experience in computer vision, used the early Cognex machines and more primitive ones from Advanced Imaging (now defunct), but it was all back in the late 80's and early 90's. The good news is the equipment has matured for standardized functions like finding alignment targets or measuring features, and automation integration is easier if you use these integrated solutions (Cognex, SICK, etc.),

It's not cheap, but the high cost of the standard equipment offsets the high cost of developing custom applications.

Give a good description of your vision application -- what do you need it to do -- and maybe someone here with more recent experience can help you with specific recommendations.

 

.

Give a good description of your vision application -- what do you need it to do -- and maybe someone here with more recent experience can help you with specific recommendations.

He is apparently sorting buckets of dirty bolts with an undetermined number of bolt types, sizes, thread pitch and grade. Also unspecified is the amount if dirt on each bolt.
http://forum.allaboutcircuits.com/threads/automated-bolt-separator.116124/

It's not cheap, but the high cost of the standard equipment offsets the high cost of developing custom applications.

Those must be some really valuable bolts. Maybe he is sorting all the fasteners of all the appliances in all of the houses that I every lived in. Real collectors items.

 

http://www.keyence.com/products/vision/vision-sys/index.jsp

 

In a nutshell, with some caveats, I don't think you'll have much success with a vision system.

Randomly oriented bolts on some kind of conveyor can be measured by vision but will require a specialized application. Then you'd need a lot of smarts in the so-called "PLC" to receive locations and sizes of bolts and some kind of robotic means to do the sorting.

If they're oriented and aligned in a 1-bolt-width conveyor so their long dimension is always pointing the same way it'd be a lot easier, but would still take a lot of programming. And you'd still need some kind of actuator at the end to do the sorting.

One of the standard vision companies may have a bolt-recognizing algorithm already done if you can singulate them (orient and separate them). So the vision might be easy if they've already done it. Otherwise, it'll be a big algorithm design and programming project.

And you still have the sorting issue. I think your major challenge is going to be mechanical, with front-end (singulation) and back-end (sorting) mechanical engineering and automation effort. How are you ever going to sort them?

 

I worked on a project in the 90's detecting blemishes on radishes for a radish farm. it had a 10' wide metal chain belt and two linescan cameras (top and bottom) forming a continuous image, which we processed in a customized pipelined processor that we designed and built. I did all the vision and automation programming. Our sorting was done with air jets at the ends blowing radishes off the belts onto a "bad" conveyor. The radishes were singulated into lanes at the beginning of the belt. The belt ran fast -- we processed 40 radishes a second.

We only had to sort good and bad, not different sizes.

The machine cost a quarter-million dollars. Probably half of that was for the customized electronic processing elements. The other half was all the mechanics.

 

You should implement an Avatar into your visual system.