Image analysis

The image analysis feature generates tags relevant to the content of images like people, animals, objects, scenery, or actions.

This feature relies on Azure Cognitive Services, a tool for image recognition. Sitecore DAM^TM integrates it to generate asset auto-tags, descriptions and color recognition data (on images and vectors only).

The following workflow represents the image analysis flow in Sitecore DAM^TM:

You can find the Image analysis section on the Asset Detail page:

The Vision media processing flow handles the image analysis feature. The flow can be found under Manage > Media processing.

The image analysis feature supports the following file types:

• Images: bmp, bpg, cr, cr2, crw, dib, dng, dpx, erf, gif, jfif, jp2, jpe, jpeg, jpg, jxr, nef, orf, raf, raw, srw, tga, webp
• Vectors: ai, eps, png, psb, psd, svg, tif, tiff

Setting up image analysis

To set up image analysis, you need to:

1. Set up the M.AssetType taxonomy

2. Enable the Vision service

Set up the M.AssetType taxonomy

The image analysis feature is applied only to assets having specific asset types. You can choose which asset types trigger the image analysis feature by configuring the M.AssetType taxonomy. You can select existing asset types or create new ones.

To add a new asset type to the M.AssetType taxonomy:

1. Go to Manage > Taxonomy:

   

2. Find and select the M.AssetType taxonomy:

   

3. Click to add a new M.AssetType taxonomy item:

   

4. Enter a Label. In this example, the Label is Image Analysis. The Identifier is auto-generated:

   

5. Click Save.

Enable the Vision service

After identifying the asset types that should trigger image analysis, you need to enable the Vision service on their associated taxonomy items.

To enable the Vision service on an M.AssetType taxonomy item:

1. Go to Manage > Taxonomy.

2. Find and select the M.AssetType taxonomy.

3. Find the taxonomy item you need to configure and click to edit it:

   

4. Set Triggers vision to True:

   

5. Click Save.

Applying image analysis

To apply image analysis to an asset, you need to assign the asset type having Triggers vision enabled to the asset.

You can set up your assets to use this in several ways:

• Apply image analysis to a single asset
• Apply image analysis to several assets using edit in table
• Apply image analysis to several assets using mass edit

Apply image analysis to a single asset

1. Go to the Asset Detail page and click in the Overview section:

   

2. In the Type field, click :

   

3. Select the asset type with Triggers vision enabled, and click OK.

4. Back in Overview, click Save:

   

5. After saving, your asset gets analyzed by Azure Cognitive Services. You can see this in the job details, found on the Background processes page:

   

6. You can see the results in the Image analysis section on the Asset Detail page.

Apply image analysis using edit in table

1. On the Assets page, select your assets and click . An editable table is displayed

2. Double-click the Type cell of the first asset, select the asset type having Triggers vision enabled, and click OK

3. Drag the green box to fill in the Type cells on the remaining assets.

4. Click Save and close.

5. After saving, your assets get analyzed by Azure Cognitive Services. You can see the results in the Image analysis section on the Asset Detail page.

Apply image analysis using mass edit

1. On the Assets page, select your assets, click , then select Edit to mass edit your assets

2. On the Assets mass edit page, click to update the Type field

3. Select the asset type having Triggers vision enabled, and click OK

4. The new asset type is set by default to replace any existing value. Click Apply to save your changes

5. Click Confirm to confirm your changes.

6. After saving, your assets get analyzed by Azure Cognitive Services. You can see the results in the Image analysis section on the Asset Detail page.