Analyzing picture taken by camera on robot via python app

    Greetings,

    I would like to analyze pictures taken by a camera on my Fanuc robot. I want to make that a loop process. So every picture that the camera takes needs to be sent to a python app and then the app should work with the picture and when it's done robot needs to take another picture and repeat the process. Any advice? I was thinking of some kind of TCP/IP communication via Karel? Since I didn't work with IR vision before can you give me some advice? Where to start (some example, manual, what functions in Karel do I need to use, etc.)?

    Thank You!

    Quote from Damek123

    .. taken by a camera..

    What kind/brand of camera?

    There exist integrated solutions from Fanuc, where you don't have to analyze anything by an external python script.

    If you use a standard camera, look for the instructions for that camera, there is nothing special with using standard cameras with Fanuc robots. You can connect the camera direct to your computer where the script is running.

    You only need an interface to the robot to know when robot reached the inspection position.

    Well, it doesn't have to be python. I said that because I was thinking that was the easiest way. What I want to do is to try to read the serial number from a cylinder and convert it to a string. Is that possible in IR vision? Do You maybe have some examples? Do You have advice about what functions do I need to use? Do I need to use Karel? Any advice on what's the easiest way to do that?

    I will use a standard Fanuc industrial camera, nothing special. I just want some guidelines on how to start.

    Thanks!

    Is the serial number printed or is it a barcode? iRVision has barcode reading. I don't know if it has optical character recognition.

    Below are 3 options. #1 and #2 are more industrial solutions.

    1. iRVision and a Fanuc camera to read the serial number. You would have to find out what iRVision package is capable of reading the serial number.

    2. Use a third party vision sensor such as Cognex or Keyence. They have great tools for barcodes and OCR. You would need to check what protocol is best for you to receive the serial number string from the sensor. The cost will be similar if you have to buy any Fanuc software packages.

    3. Build your own setup with a PC, GigE camera, and custom code such as Python. You would probably use TCP/IP to send commands and to receive the serial number string.

Advertising from our partners