El Bruno

Hi !

In the last couple of days, a new set of services were added to Microsoft Flow. In the Microsoft Flow Blog, they shared a complete description of the new services. I’ll focus this post on Computer Vision actions (part of Microsoft Cognitive Services suite). We will be able to perform image analysis, text recognition and more.

So, let’s try this with a very simple flow:

• This Flow will start when a new file (image) is added to a OneDrive folder
• The content of the image will be sent to Computer Vision to analyze the content
• The result of the analysis process will be sent as a mobile notification

The next image displays a running log of this Flow.  The image saved on OneDrive display a group of people. The results are very accurate.

The parameters to be used are very straight forward and we can easily match a OneDrive file with a Computer Vision action using the Analyze Image action definition.

In the Computer Vision actions we can use the following ones:

These actions are the same ones we can find in the API definition for Computer Vision (link). The extra one is related to “Recognize Domain Specific Content”. Which is a very interesting one.

Greetings @ Toronto

El Bruno

References

• Flow Blog, Build flows with HTML Tables, CSVs and other new features like OCR and image content detection
• El Bruno
  • How to invoke a Flow from an external app using HTTP Request
  • How to review a Flow process
  • Push Notification Support
  • Calling a Flow from a C# ConsoleApp
  • Support for Team coauthoring and sharing
  • Email Sentiment Analysis on a PowerBI dashboard, and Flow
  • Schedule a Flow execution, recurrence supported!
  • How To iterate through collections of elements (Excel, Sharepoint, Cognitive Services and more…)
  • “Save as ” and interaction with physical buttons FLIC and BTTN