logo
logo

Image analysis

The image analysis feature automatically generates tags, descriptions, and color data for image assets using Microsoft Azure Cognitive Services, which is an image recognition tool.

Important

Microsoft imposes limits and quotas on the images you can send to Azure Cognitive Services.

The Image analysis section is located on the assets details page, as shown in the following example:

To use the image analysis feature, it must be set up in your solution. You must also edit the asset and assign it an asset type that has Triggers vision enabled. In the example, the asset type is Social Media Asset, which is a defined taxonomy.

The image analysis feature supports the following file types.

Asset typeSupported files
Imagesbmp, bpg, cr, cr2, crw, dib, dng, dpx, erf, gif, jfif, jp2, jpe, jpeg, jpg, jxr, nef, orf, raf, raw, srw, tga, webp
Vectorsai, eps, png, psb, psd, svg, tif, tiff
Note

The image analysis feature does not process video assets, as they have their own flow.

Limitations:

  • You can only edit the Tags and Colors properties.
  • You cannot train the AI.
  • You cannot improve the results but you can change the confidence level in the vision media processing flows, which is by default set to 0.5.

Can we improve this article ? Provide feedback