Are there any serious implementations of vision systems (i.e. OCR) in Ignition?
Are you expecting Ignition to do the OCR analysis? Typically the camera will have that functionality and do the recognition work, then spit the resulting string out into an interface accessible by ignition (TCPIP Sockets, OPCUA, or other standard interface).
Thanks for the response Oscar.
I’ve got snapshots of bitmaps being packaged in a password protected webpage to convey a small collection of remote SCADA data.
I’ve pointed a native Ignition IP camera component at the changing *.bmp file URL for a ‘realtime’ display, but
historizing these numbers will require some form of OCR, as I understand it.
I’m guessing I’m not alone, but perhaps there isn’t a very big demand for such for reasons you’ve pointed out.
Thanks for the explanation, I do see where you are going with this.
Ignition can certainly do what you need it to do, but I would recommend at looking into other avenues. Specifically, I would consider leveraging cloud services such as Azure or AWS to offload the image processing outside of your Ignition Gateway.
There are already tools within these platforms to do OCR; one example is Amazon Rekognition (https://docs.aws.amazon.com/rekognition/latest/dg/text-detection.html). You can interface from Ignition using their API endpoints and get your data into historian (if historian is were you need them to go ultimately)
Exciting project! Keep me posted on the progress.
Thanks for the tip Oscar. I’ll look into their pricing.
Displaying this data is already a bit slow in the current IP camera solution, particularly on initial startup, so I would have to roll my sleeves up to figure out how to get the server to do a lot of the pre-processing anyway to keep the data flowing to the client as streamlined as practicable.