Azure’s Computer Vision service provides developers with access to advanced algorithms that process images and return information, depending on the visual features required.

Prerequisites for MSVision Node

1. Create a Microsoft Azure account and navigate to Portal.
2. Generate the Subscription Key and Service URL:
   1. Create a Resource Group.
      
   2. Add resources to the created Resource Group.
      Click the Resource Group and click the + Add button.
   3. From the AI + Machine Learning section, add Computer Vision to the Resource Group.
   4. From the Keys and Endpoint section of the added resource, note either Key 1 or Key 2 as Subscription Key and Endpoint as Service URL.
3. In your machine, install the required packages to execute the MSVision Functions. To install the packages:
   1. Open the command prompt.
   2. Execute the following command to change the directory to C:\jiffyservice\precompile\connector\requirements.

      cd C:\jiffyservice\precompile\connector\requirements

   3. Execute the following command to install the packages.

      pip install -r msvision.txt –user

If the command prompt is not recognizing pip, use the command C:\Python37\Scripts\pip install -r msvision.txt.

Design the Action steps for MSVision Node

In the Actions tab, you can define actions to be performed in MSVision Processing Node.

Click the MSVision node to open the attributes pane. In the Actions tab, select one of the following options from the first drop-down:

• MSVision Functions: To select the required MSVision Functions from the list in the second drop-down.
• Custom MSVision Function: The Custom MSVision functions are created for specific needs if not met by the Predefined MSVision functions. All the User Defined Functions created in the HyperApp with Type selected as MSVision gets listed in the drop-down.
• Call Function: To select other generic Functions from the list in the second drop-down.
  To know more about how to use Functions, click here.
• Set Variable: To set the value for a Variable as needed in further steps. The value set can either be a constant or another variable. If this variable is toggled as an output variable in the Variables tab, it can be mapped to succeeding nodes of the task.

Variables

All Input/Local variables of the MSVision node are displayed under the Variables tab.

Table Definition

All Default/Custom DataTables in the task are displayed under the Table Definitions tab.

Properties of the MSVision Node

1. Navigate to Properties and enter Name and Description.
2. Enable Mark run Failure on Node Fail and Continue on Failure? options as required.
3. Select the Cluster to run on the task from the Clusters drop-down.

Result of Execution in MSVision Node

After executing the task, the Result of Execution window displays the following tabs:

• Input: The inputs mapped to the MSVision node are displayed.
• Output: The results of the actions performed are displayed in XML format.
• Run Info: The Run Details and Machine Details are displayed.
  • Run Details include Run ID, Sequence Number, and Total Node Execution Time (in seconds).
  • Machine Details include ID, Name, Host, Port, Iteration Start Time, Iteration End Time, Iteration Time (in seconds).
• Step Data: The details of the steps designed inside the node, the status of each step, and the values stored in the variables are displayed.
• Formatted Output: The output is displayed in the table format.

When Get Local Image Colors(123456abcs, https: //jiffyai.azure, C:\Invoice.jpg) function is executed, dominant colors in the image are fetched.