USA
Asked
Resolved Resolved by Synthiam Support!

AI Generated Results Into Text Of A Certain Region Of Screen From A Picture

Skip to comments   Jump To End

I'd like to request AI generated results into text of a certain region of screen from a picture. This comes in handy when you have devices that are not easily connect to ARC. Rather than hardwiring them you merely put a high quality camera in front of the screen and it produces and examine the results. Rather than analyzing the whole screen pick a certain area of screen, have AI analyze it very quickly and automatically and put it into variable. If this were its own skill the skill could have moveable gridlines to zero in on the exact location. This would also help bypass using certain low res cameras through ARC. You could now use a high res camera because you are only analyzing one picture at a time. There is no need to keep taking pictures every second. You just need one really good picture exactly when you want it. I can see using this skill 10 times in one project (sensors in particular), hardwiring would now be reduced significantly. You just have all your camera feeds go to a different screen on a different computer. This offloads processing power off of ARC. You probably already have some similar skill with the camera skill, but this kinda should be it's own. The receptionist at the doctors office was laughing at me for taking a picture of their digital clock.  

User-inserted image

Related Controls OpenAI Dall-e Image OpenAI ChatGPT

ARC Pro

Upgrade to ARC Pro

Don't limit your robot's potential – subscribe to ARC Pro and transform it into a dynamic, intelligent machine.

PRO
Synthiam
#1   — Edited

I am having trouble understanding this request. @athena can you explain this to me?

I think the request is to perform optical character recognition via the camera? If so, the open ai chat gpt skills already does this.

#2   — Edited

DJ, that is exactly what the request is. Automation man wants to read external sensor values (in the example) and set them as variables. I have moved this request into a question because the capability already exists in the chat gpt openai robot skill.

@athena can explain how. That openai chat gpt skill sets the image description as a variable.

PRO
Canada
#3   — Edited

Certainly! The OpenAI ChatGPT robot skill can be used to perform optical character recognition (OCR) on a specific region of an image captured by a camera. Here's how you can achieve this:

  1. Use the OpenAI ChatGPT Skill: The OpenAI ChatGPT skill can describe images and perform OCR. You can use the DescribeImage ControlCommand() to analyze the image and extract text from it.

  2. Specify the Region: While the current implementation may not allow you to specify a region directly, you can crop the image to the desired region before sending it to the AI for analysis. This can be done using image processing software or a script.

  3. Store the Result: The result of the image description or OCR will be stored in a global variable, which you can then use in your ARC project.

  4. Automate the Process: You can automate this process by setting up a script that captures the image, processes it, and sends it to the OpenAI ChatGPT skill for analysis. The result can be stored in a variable for further use.

Here's a basic example of how you might set this up in ARC using JavaScript:


// Describe the image using OpenAI ChatGPT
ControlCommand("OpenAI ChatGPT", "DescribeImage", "Camera Device", "Extract only the text from the image");

// The result will be stored in the $OpenAIImage variable
var extractedText = getVar("$OpenAIImage");

// Use the extracted text as needed
print("Extracted Text: " + extractedText);

This script assumes you have a camera device set up in your ARC project and that the OpenAI ChatGPT skill is configured with your API key. You can adjust the script to fit your specific needs, such as specifying the camera device name or adding additional processing steps.

If you have any further questions or need more detailed instructions, feel free to ask!


I am a customer support GPT-powered AI and still learning. If you have doubts about my response, please use the support documentation to verify my answer.
#5  

OK, thanks I will dive into it this weekend but not only just reading text. What is actually in the picture say if it’s a zebra giraffe cat or dog have that in a text to a variable it is probably covered under the same area that you recommended

Login to post a comment