The Zone intrusion detection parameters. The Administrator client

Frames per second.

fps

float

10.0

[1, 30]

A certain number of frames per second a video is processed.

Prediction confidence threshold.

confidence_threshold

float

0.534488676

(0, 1)

Detection score threshold is applied for predicted objects.

Motion detection history size.

mask_history_size

int

3

[1, 10]

Number of motion masks used for averaging. This setting helps to reduce false positive detections caused by adverse weather conditions (like rain or snow). A higher value can decrease false positives but may also lead to the suppression of small or dynamic true positive detections.

Minimum time before generating a repeat event.

event_cooldown

float

1.0

(0, +inf)

Delay between similar events in seconds.

Motion area expansion.

motion_margin

float

0.5

(0, +inf)

Additional context that will be taken around the motion area before sending it to the detector. Denotes the size relative to the rectangular motion size. Thus, 0 means no padding, 1 – increasing the motion area by two sizes of the area, and so on.

Intersection threshold of detected objects.

nms_iou_threshold

float

0.5

(0, 1)

Overlapping ratio between two boxes used to consider the boxes as a single detection. Increased value leads to more duplicates. Decreased value may lead to skipping close detections.

Dense history time interval.

dense_history_interval

float

0.808510686

(0.1, +inf)

Time interval (in seconds) that is used to keep frames in history. The oldest frame is used to compare with new frame.

Sparse history of time interval.

sparse_history_interval

float

300

(1., +inf)

Временной интервал (в секундах), который используется для сохранения разреженных кадров в истории. Логически он должен быть больше, чем dense_history_interval.

Time interval (in seconds) that is used to keep sparse frames in history. Logically, should be greater than Dense history time interval.

Update frequency of sparse history.

sparse_population_frequency

float

60

(1., +inf)

Frequency (in seconds) with wich we populate sparse history. Thus, sparse_history_interval=300 and sparse_population_frequency=60 means that we keep 1 frame per minute for 5 minutes (5 frames).

Immobility threshold

still_threshold

float

0.5

(0, 1)

A threshold used to eliminate false positive motion detections when an overlapped stationary object becomes visible again. The more the value the more previous frames the object should be present in to be considered as still object.

Sensitivity

sensitivity

float

0.5

(0, 1)

Signal magnitude threshold for creating an event.

The greater the value, the more events we get (and more false positives) and vice versa.

A minimum object size.

min_size

float

0.1

[0, 1]

A minimum relative size of a moving object required for a successful detection.

A maximum object size.

max_size

float

0.7

[0, 1]

A maximum relative size of a moving object required for a successful detection.

Immobility check

still_check

bool

True

{False, True}

Whether check still objects.

Initialization duration.

init_duration

float

5.0

[1, 10]

Initialization duration.