Page History

...

1. A face is dynamically framed on the frame;
2. If Face Intellect is installed, then there is possibility to record a frame with a face to the file of .bmp format to the folder «<Face-Intellect>\ Bmp\ person\».

To configure Face detection tool, do the following:

1. Go to the Hardware tab in the System settings dialog box (1).
2. Select the Face detection object corresponding to the required detection tool, on the Hardware tab in the objects tree (2). The settings panel for the selected object appears on the right of the Hardware tab.
   
3. Specify how many fps from the video stream are processed by the Face detection program module. This parameter is necessary for reducing the load on the Server (3).
4. Specify the size of a face in percent relatively to the sizes of camera’s video surveillance window when the face is required to be dynamically framed (4).
5. If the color of a dynamic frame is to be changed, then left click the Color field (5). Select the required color in the standard Color dialog box and click OK.

6. Disabling of Process color video function (6) is only feasible if Face Intellect is installed. It is not recommended to disable it.

   Info
   title Note.
   If the Process color video checkbox is unset, the color video image will be converted to black and white image to generate the vectors by the black-and-white frames. This function is supported by Tevian 2.3.0 face recognition module only. It reduces the CPU load, insignificantly affecting face capture and recognition quality (see Face Intellect Administrator's Guide for details).

   
   

7. By default, the Face detection tool processes a video with the resolution set for the Camera parent object.  To process a video with 320x240 pixels resolution, set the Process video in low resolution checkbox (7).

   Info
   title Note.
   Enabling this option reduces the load of the Server’s processor, but the probability of face detection also reduces.

   
   

   Note
   title Attention!
   The Process video in low resolution setting is not relevant in case, when a high resolution is set for a Camera parent object.

   
   

8. Set approximate width of face on the video image in percentage of width of framepixels;

   1. In the From field enter minimally allowable width of face on the video image (48).

   2. In the To field enter maximally allowable width of face  face on the video image (59).

9. Set approximate height of face on the video image in percentage of height of framepixels;

   1. In the From field enter minimally allowable height of face on the video image (610).

   2. In the To field enter maximally allowable height of face  face on the video image (711).

      Info
      title Note.
      Faces won't be detected if specified minimal sizes highly differ from actual sizes. For example, if there are big faces (15-30%) on the video and minimal size is 5% and maximal size is 80%, then faces won't be detected. But if minimal size is 12% and maximal size is 80%, then faces will be detected. So, detection won't detect big faces if it is configured for detection of small faces, and vice versa.

   By default the Face detection tool processes a video with the resolution, set for the Camera parent object.  To process a video with 320x240 pixels resolution, set the Process video in low resolution checkbox (8)

   Estimate the part of detected faces while processing the video with 320x240 pixels resolution. In case, when this value is not high enough, set the Process video in low resolution checkbox unchecked

   1. 
      

      Info
      title Note.

   Enabling this option reduces the load of the Server’s processor, but the probability of face detection also reduces.

   Note
   title Attention!
   The Process video in low resolution setting is not relevant in case, when a high resolution is set for a Camera parent object.

10. Specify the size of a face in percent  relatively to the sizes of camera’s video surveillance window when the face is required to be dynamically framed (9)
11. If the color of a dynamic frame is to be changed, then left click the Color field (10). Select the required color in the standard Color dialog box and click OK.

12. 1. If all face sizes are set to 0, the Face Detection tool uses similar parameters set on the Face Recognition Server object settings panel being part of the Face Intellect software.

       
       

13. To pixelate the part of video image in which the face was recognized, do the following:
    
    1. To enable the pixelate function set the Pixelate area checkbox (
11
14. 1. 12).

    2. Set the Grain size cursor in position corresponding to the pixilation grain size in pixels (

12

15. 1. 13).

16. Set the Face capture area’s scaling slider to the position corresponding to the required scaling of a face in the frame for recording (1314). The current scale value is displayed in the field above the slider and ranges from 1 (only the face is recorded) to 5 (the whole frame is recorded).

    Info
    title Note.
    This setting is relevant in case of installed Face-Intellect system. A frame with a face is recorded to the < Face-Intellect>\ Bmp\ person\ folder.

    
    

17. Click the Apply button (1415).

Face detection tool is now configured.