Extensions

Overview

Description

• This extension turns text from OBS sources into SAMMI variables either with Google Vision or Iron OCR

Google Vision Text Recognition Pros

• Accuracy: Google Vision OCR is powered by Google’s advanced machine learning algorithms, making it highly accurate in recognizing text from images.
• Multiple Language Support: Google Vision OCR supports a wide range of languages, making it suitable for international applications.
• Regular Updates: Google continually updates its services to improve accuracy and add new features.
• Multiple Requests can be made at the same time
• Can also be used to output an object that contains all texts in an object array with their coordinates
• New users can 300$ free credits for testing in the first 3 months

Google Vision Text Recognition Cons

• Cost: You will be billed by usage after your 1000 free Text Recognitions per month were used (currently you pay 1,50$ per 1000 additional images)
• Dependency on Internet Connection: It requires a stable internet connection to access the API, which can be a drawback if you need to work offline.
• Privacy Concerns: Uploading sensitive data to a cloud service may raise privacy and security concerns, especially for confidential documents.
• Rate Limits: Google imposes rate limits on API usage, which could affect your application’s responsiveness during high-demand periods.

IronOCR Pros

• Offline Processing: IronOCR allows you to perform OCR locally, making it suitable for applications that need to work offline or have data privacy concerns.
• Cost Savings: Once you purchase the IronOCR library, there are no ongoing costs or API usage fees, making it cost-effective in the long term.
• Speed: Local processing can be faster than making API calls, especially for large volumes of documents, as you don’t have to deal with internet latency.
• Trial licenses for 30 days can be generated per mail

IronOCR Pros

• Initial Setup: Setting up and configuring IronOCR is more complicated
• Accuracy: While IronOCR is generally accurate, it may not match the accuracy of cloud-based services like Google Vision OCR, which benefit from continuous machine learning improvements.
• Language Support: IronOCR may have limitations in terms of language support compared to cloud-based services with extensive language libraries.
• Resource Usage: OCR processing can be resource-intensive, and running it locally might require powerful hardware, especially for high-throughput applications.
• Multiple requests queue up instead of running async