When you click a camera's Setup icon in the FD sub-menu the FD Setup screen will be displayed.
In the FD setup screen you can select an area of the camera's image you want to monitor for people's faces.
Selecting Your Camera
You can select the camera you want to configure the detection area for from Camera drop down list.
Creating A Detection Area
Creating a detection area allows you to specify a portion of the camera's image to monitor rather than the entire screen. You have two options when creating a detection area, you can define a box and monitor within it or draw a line a monitor as people cross over it.
Face Detection Within A Specified Box
To create a boxed detection area follow these easy steps:
1.Set the Rule Kind drop down list to Rect.
2.Set the Detection Area drop down list to Customise.
3.Click on the camera's image where you want to set the top left corner of the detection area.
4.Click where you want to set the bottom left corner.
5.Click where you want to set the bottom right corner.
6.Finally click where you want to set the top right corner of the detection area.
You will notice as you are setting your corner points the recorder will draw a red box between between the points.
Face Detection When Crossing A Line
You can set up face detection to only trigger when somebody crosses over a line to better control when faces are captured. You can even set which direction the line is crossed before a capture so you can monitor for face on entry or as people exit. To set up line crossing face detection follow these simple steps:
1.Set the Rule Kind drop down list to Line.
2.Click on the camera's image where you want to start the line crossing.
3.Click where you want to stop the line crossing.
4.Finally select which direction a person crosses the line before face detection is triggered from the Rule Type drop down list, from side A to B (A->B) or from side B to A (B->A).
You will notice as you are setting your points the recorder will draw a yellow line between them.
Editing A Detection Area
You can edit the existing detection area by simply dragging any of the red corner squares to where they are required.
Determining When Thumbnail Snapshots Are Captured
You can determine how the recorder will capture thumbnail snapshots when a face is detected using the Snap Mode drop down list.
Default: Will only capture a snapshot when the face are first detected. Additional thumbnail snapshots will only be captured if the person leaves the camera's image and returns.
Realtime: Will only capture a snapshot from just before the face leaves the image.
Interval Mode: Will allow you to specify how many and how often thumbnail snapshots are captured using the Snap Num and Snap Frequency fields.
Number Of Snapshots
You can set how many snapshots are captured for each face using the Snap Num field.
1: Will capture a single snapshot when the face is first detected.
2: Will capture a snapshot when the face is first detected and 1 further snapshot after the period of time specified in the Snap Frequency field if the face is still present.
3: Will capture a snapshot when the face is first detected and 2 further snapshots, 1 each time the duration specified in the Snap Frequency field elapses and the face is still present.
Unlimited: Will continuously capture snapshots each time the duration specified in the Snap Frequency field elapses and the face is still present.
Snapshot Frequency
You can set how often the recorder will capture a snapshot of each face using the Snap Frequency field. Enter the amount of time (in seconds) you want between each snapshot to set the snap frequency.
This timer will remain active as long as the face is present. If the person leaves the camera's image and returns a snapshot will be captured and the timer will start over.
Minimum Size Of Objects Detected
You can set the minimum size of objects that are detected in the Min Pixel field. You can enter a minimum width between 64 pixels to 1080 pixels.
This is useful as the area you are monitoring will differ from camera to camera. If the area you are monitor is far from the camera you you would want to set a small minimum width as pedestrians and vehicles will appear small on-screen. However, if the area you are monitoring is fairly close to the camera so that pedestrians and vehicles appear quite large you can set a higher minimum width to avoid false triggers.
Mug Shots Or Any Angle
For the most accurate face detection a straight on mug shot is usually the best option but the Zip AI cameras are also capable of detecting a face from most angles.
Depending on the installation a straight on shot may not be possible. You can choose your viewing angle from the Apply Mode drop down list as described below:
Frontal View
Great for entrances and corridors Frontal View optimises the camera for straight on face detection.
Multi Angle
Ideal for general surveillance Multi Angle optimises the camera so that it can also detect faces at various angles.
Customise
Selecting customise unlocks 4 extra settings so that you set up bespoke detection angles to suit your requirements as described below:
•Roll Range - is how far somebody can tilt their head (ear towards their shoulder) and still be detected as a face. You can set the angle from 0 degrees so that a face would have to be perfectly straight to 180 degrees so that the face can be tilted 90 degrees to either side.
•Pitch Range - is how far a person can look up or down and still be detected. At 0 degrees the subject would have to be looking directly at the camera. At 180 degrees the subject can be looking towards the ground or towards the sky and could still be detected.
•Yaw Range - is how far somebody can be looking to the side and still be detected. At 0 degrees the subject would have to be looking directly at the camera. At 180 degrees the subject can be side on to the camera and could still be detected.
Snap Frequency
You can set how often the recorder will capture a snapshot of the subject using the Snap Frequency field. Enter the amount of time (in seconds) you want between each snapshot to set the snap frequency.
This timer will remain active as long as the subject is present. If the subject leaves the camera's image and returns a snapshot will be captured and the timer will start over.
Minimum Size Of Objects Detected
You can set the minimum size of objects that are detected in the Min Pixel field. You can enter a minimum width between 64 pixels to 1080 pixels.
This is useful as the area you are monitoring will differ from camera to camera. If the area you are monitor is far from the camera you you would want to set a small minimum width as pedestrians and vehicles will appear small on-screen. However, if the area you are monitoring is fairly close to the camera so that pedestrians and vehicles appear quite large you can set a higher minimum width to avoid false triggers.
Detection Sensitivity
By setting the Sensitivity you can avoid false triggers or capture slight movements. You can enter a sensitivity level between 0 and 100.
0 is the lowest sensitivity level and will only trigger when the camera gets a perfect, clear view of the subject so that the recorder can tell with a high degree of accuracy that the object is a pedestrian or vehicle. This will avoid false triggers but could miss triggers as the object was in the ideal position or in the best lighting.
100 is the highest sensitivity level and will trigger when the recorder thinks that any object could possibly be a pedestrian or vehicle. This is handy when the camera may only see the subject for a brief moment or the view may be partially obscured. However, this is a trade off as setting a high sensitivity could result in false triggers.
Pedestrians, Vehicles Or Both
By setting the Detection Type you can choose what the recorder looks for in the camera's image. You can tell the recorder to look for pedestrians, vehicles or both by ticking or unticking the Pedestrian or Vehicle check boxes.
Stationary Pedestrians And Vehicles
The Detection Mode drop down list allows you to monitor for stationary subjects as well as moving ones.
By default it will be set to Motion Mode which will only detect moving pedestrians and vehicles. This is ideal in areas where there might be parked cars so that they don't continue to trigger when they are stationary.
In some circumstances you may wish to also capture stationary pedestrians and vehicles such as no loitering or no parking areas. If this is the case you can select Static Mode from the drop down list. The recorder will still trigger when moving subjects are detected but will now capture stationary ones as well.
Full Screen Monitoring
You can set a detection area as described previously but in some cases you may just want to monitor the entire image. You can do this using the Detection Area drop down list. Customize will only monitor within the defined detection area but if you select Full Screen the recorder will monitor the entire image.
Saving Changes
When you are happy with the current settings click the Save button. If you exit the window without saving any changes will be lost.