| |
Detection
with IR illumination
A
pair of thieves detected in the act. |
|
| |
|
 |
Virtual
Fence Alert
Alerts
can be triggered when a virtual tripwire
is touched or crossed.
|
|
| |
|
 |
Fenceline
Detection within Moving Leaves and Shadows
“Fence
line detection in challenging conditions”
; Video security experts know that positive
detection within an Area of Interest which
is experiencing environmental noise (e.g.,
moving leaves and their shadows) is one
the most difficult scenarios for any analytics
system, at any price. Amazingly, the WiseEye
meets and even exceeds the performance level
of the highest-end analytics costing 3-5X
more. |
|
| |
|
 |
Construction
site: Fast -moving intruder
Fast-moving
detection: construction site . Many analytic-based
systems go through a series of internal
processes before a detection decision/notification
is made. Unfortunately, some events (such
as running people) make clear object-recognition
difficult, and the analytics may decide
to ignore the event altogether. As a result,
these “false negatives” put
the entire site at risk. |
|
| |
|
 |
Fenceline
Detection: Crawling Intruders at a Distance
Sterile
zones along fence lines are a mandatory
requirement for many perimeter protection
applications. Detecting crawling intruders
close to the fence is a difficult task for
all analytic solutions--especially at distance.
A crawling intruder’s small size and
slow movement speed at 300 feet makes detection
very difficult. The WiseEye sensor, with
its built in sterile zone function, accurately
detects the crawling intruder (it ignores
any activity outside this zone) and also
tracks him when leaving the sterile zone. |
|
| |
|
 |
Positive
Detection with Wind and Shadows
The
“masking” function is often
used by analytics products to prevent false
positives. Unfortunately, “noisy areas”
are not always under our control, and in
many cases intruders must be detected within
that very area—rendering masking useless.
In this example, the WiseEye’s strong
filtering of background changes ignores
the “noise” (generated by trees
and shadows) while still detecting a real
moving in the same area is detected. |
|
| |
|
 |
250
ft. Night Detection with IR Illumination
WiseEyes
can be successfully used for long range
detection even in totally dark conditions.
The monitoring operator can make a “classification”
level assessment (i.e., human, not deer
or garbage can, etc.) to speed real-time
response.
|
|
| |
|
 |
People
and car lights
The
WiseEye can make positive detection decisions
in changing lighting conditions (the passing
car lights). People are still tracked under
these difficult scenarios. |
|
| |
|
 |
Four
WiseEyes monitoring a water facility
Typically,
several WiseEyes are used to insure complete
visual protection for the entire facility.
Here, images from four sensors are shown
simultaneously. One sensor (top left) was
mounted on top of the water tank, the other
three on the surrounding fence posts. Detection
results are shown from a few hours of system
testing by the facility's security officers.
Tests started at 5AM (before sunlight) and
ended at 8AM (in full sunlight). No false
alarms were produced even though one sensor
(on tank top) was pointing exactly at the
rising sun. |
|
| |
|
 |
Detection/False
Positive Suppression in Changing Weather
Unlike
other analytics-based systems, WiseEyes
automatically adapt to year-round weather
conditions, and one of the most significant
challenges for video detection cameras in
the 100-yard application is an abrupt weather
change. In this example, a 3-inch snow shower
passes thru the Area of Interest. The sudden
introduction of larger, slow-moving snowflakes
contrasts dramatically with sunny conditions
an hour earlier. The WiseEye’s automatic,
adaptive analytics algorithms ignore the
snow—while still detecting the vehicle.
The key element here is no configuration
change was required on the WiseEye to handle
this weather pattern.
|
| |
 |
False
Positive Suppression at a Pumping Location
Weather
and illumination can change drastically
even in a short time. In this compressed
clip we see snapshots of severe illumination
changes during a 24 hour period in a Washington
State energy facility. No false alerts were
generated by the WiseEye security system. |
|
| |
|
 |
Filtering
Tree Movement without Masking
A
highway in Beijing is monitored. Note that
tree movement does not generate alerts.
This is recognized automatically by the
algorithm without need for manual masking. |
|
| |
|
 |
New
Mexico burglary attempt thwarted with the
help of a WiseEye security system. |
|
| |
|
 |
Water
Utility Protection
Detection
results from an installation near the Israeli
border. The WiseEye watches the scene and
only transmits alert frames upon a positive
detection. |
|
|
|
|