Clothing Analysis Results

Clothing analysis can produce several results for each person. When you set ClothingMode=Full,Upper,Lower, Media Server can produce up to three results for each person, describing the clothing covering the full body, upper body, and lower body. Media Server does not produce a result if a region is outside the image, or if clothing analysis cannot find the region.

The following XML shows a single record produced by clothing analysis.

<record>
   ...
   <trackname>Clothing.Result</trackname>
   <ClothingResult>
     <id>63f77e13-e916-4528-acfb-559e1db28574</id>
     <parentID>0fba065c-48d4-442b-9801-b10f84c4b3b6</parentID>
     <region>
       <left>116</left>
       <top>190</top>
       <width>248</width>
       <height>480</height>
     </region>
     <regiontype>Full</regiontype>
   </ClothingResult>
</record>

The record contains the following information:

  • The id element contains a unique identifier for each clothing result.
  • The parentID element contains an identifier that matches one or more input records. This means that if the person was identified by face detection the identifier matches the ID for the detected face, and if the person was identified by object class recognition the identifier matches the ID for the recognized person. So while all clothing analysis results have a different id, results that correspond to the same person share the same parentID.

  • The region element contains the location of the clothing. The values are in pixels. The left and top elements give the position of the top-left corner of a rectangle, and width and height provide its width and height.

  • The regiontype element specifies whether the region represents clothing covering the full body (Full), upper body (Upper), or lower body (Lower). These values correspond to those for the configuration parameter ClothingMode.