UiPath Document Understanding explanataion

Help Activities
,
1
  1. If I set degree of parallelism more than my number of cores then what will happen?
  2. In OMI OCR what is the use of by default “Image” variable in input image ?
  3. Please explain “Profile and Scale” in OMNI OCR

Thanks,
Ritaman.

2

@Ritaman_Baral

1.Parallel processing is a technique where multiple tasks are executed simultaneously by utilizing ultiple CPU cores. Each core can handle one thread or task at a time. When you set the degree of parallelism to a higher value than the number of cores, you are instructing the system to run more tasks concurrently than it has physical resources to handle. 1If the tasks are CPU-bound (i.e., they require significant processing power), running more tasks in parallel than the number of cores can slow down the overall execution because the CPU will spend more time on context switching than actual processing.

2.The “Image” variable is typically used to pass the image that contains the text you want to extract. You can provide this image in various formats, such as JPG, PNG, BMP, or other compatible image formats supported by the OCR engine. The OCR engine then analyzes the image to recognize and extract the text from it. When you use the Omni OCR activity in UiPath or any other OCR engine that supports it, you will find a parameter where you can pass the “Image” variable. Depending on the OCR engine and the activity being used, the name of the parameter may vary, but it generally has a clear indication that it expects the image to be processed.

3.Profile:

  • The “Profile” parameter allows you to choose a specific OCR profile that best suits the characteristics of the text in the input image. Different OCR profiles are designed to handle various types of content and text layouts more effectively.
  • For example, if you have an image containing printed text, you might choose a “Printed” profile. Similarly, if the image contains handwritten text, you can select a “Handwritten” profile. These profiles are optimized to deliver better results for their respective text types.
  • Some OCR engines might offer additional profiles for different languages, fonts, or document layouts, allowing you to pick the most suitable one for your specific use case.
    Scale:
  • The “Scale” parameter allows you to specify the scaling factor or size of the image that the OCR engine will process.
  • In certain scenarios, the scale of the image can impact the accuracy of the OCR results. If the image is too small, text details might be lost, leading to poor recognition results. On the other hand, if the image is too large, the OCR engine might require more processing power and time to analyze the image, affecting performance.
  • By setting an appropriate scale value, you can optimize the OCR process for your specific image size, ensuring a balance between accuracy and efficiency.

HOPE IT HELPS!!

1 Like
3

But bydefault even if i am not passing any Image . there is one Image variable.

4

@Ritaman_Baral

Yeah it will take by default only, means its expecting an image too. In new version image is the default variable.

Regards

5

@Ritaman_Baral

If you find the solution for your query please do mark as solution to close the loop.

Happy Automation!!

6

This topic was automatically closed 3 days after the last reply. New replies are no longer allowed.