6
7
• By masking an area with remarkable luminance change, it is possible to shoot images on which faces can
be recognized more accurately.
When shooting images on a place where strong light sources may affect from outside, the following settings
can reduce the light influence.
Cases not suited for face matching
The matching accuracy varies depending on installation, settings and adjustment of the camera, ambient envi-
ronment and object. Therefore, face matching may not be performed in the following cases.
• Illuminance levels are dispersed. (Example: Outdoors)
• Faces are interrupted by such objects as medical masks, sunglasses or helmets.
• Faces are not directed to the front side.
• Photographic subjects are moving quickly.
• Under the influence of strong external light (such as the headlights of a car, rising sunlight or setting sunlight)
To improve the performance of face matching
• Register the target face images from the images recorded by the camera installed under the same conditions.
• Perform settings for the camera so that the camera can shoot face images at a balanced illuminance level
both in the daytime and nighttime.
• Install the camera so that the camera can move within the assumed moving pathway of photogenic
subjects.
• Configure the settings of the people counting area and the face matching / age and gender judgment area
(
☞
Page 15) to avoid counting the same person repeatedly and avoid targeting people's faces printed on
a poster.
• Configure the setting so that the people counting area is included in the face matching / age and gender
judgment area.
<Example of the area setting>
Glass
<Example of image (Near the entrance)>
<Example of mask area setting for the image>
Poster
Around the poster Not masked Masked (White)
Glass
Poster
People counting area
Face matching / age and gender judgment area