Designing a Site with
Pelco Smart Analytics
Document number:C6717M-B
Publication date:02/23
Designing a Site with Pelco Smart Analytics
Table of Contents
Introduction 1
Classified Object Detection 1
General Guidelines 1
Mounting Height and Angle 1
Field of View 2
Object Speed 2
Camera Image Rate 3
Reflected Light 3
Headlights 3
Adaptive IR 4
Lux on Target 4
Obstructions 5
Analytics Scene Mode 5
Detection Range 6
Outdoor Areas 6
Indoor Areas 8
Self-Learning© 8
Resetting the Learning Progress 9
Camera-Specific Guidelines 9
Pelco Sarix Professional 4 Series camera 9
Pelco Fisheye camera 9
Pelco Sarix Corner Mount 3 Series camera 10
Pelco ExSite Enhanced 2 Series camera 10
Pelco Esprit Compact Series camera 10
Pelco Sarix Modular camera 10
Pelco Sarix Thermal Enhanced 4 Series camera 11
For More Information 11
Pelco Troubleshooting Contact Information 11
C6717M-B | 02/23 i
Designing a Site with Pelco Smart Analytics
Introduction
Pelco video analytics cameras are easy to install and can achieve positive analytics results without
ongoing configuration adjustments. Pelco's patented video analytics are designed to automatically adjust
to subtle changes in the camera’s field of view, like changing seasons, without configuration or
adjustment.
Objects are classified as person or vehicle. You can set up analytics events to send to a VMS in the
Camera Configuration Tool (CCT) software. CCT is available from https://www.pelco.com/camera-
configuration-tool.
For video analytics to perform effectively, the analytics cameras must be installed correctly.
Video analytics enabled cameras must be installed:
l
Within the height and angle guidelines.
l
Within sight of the area of interest.
l
Where there is sufficient light in the area of interest.
l
Where there is sufficient contrast to detect foreground motion.
For example, a person walking in white clothes in a snow-covered field of view may provide poor
results.
The following information provides a basic set of installation parameters. Read through the
entire document before installing cameras. For unique camera model requirements, see section
Camera-Specific Guidelines.
For site configurations that differ from the listed recommendations, or when in doubt, consult a
Pelco representative before installing the cameras.
Classified Object Detection
Design your site with the following guidelines to use video analytics for Object Classification.
General Guidelines
In general, cameras should be installed according to the following guidelines to achieve optimal analytics
performance.
Mounting Height and Angle
l
Cameras should be mounted at a minimum of 2.8m (9ft) level to the horizon and ground plane
(for outdoor or large indoor areas).
l
Cameras can be tilted within 30° from the horizontal for optimal object classification.
C6717M-B | 02/23 1
Designing a Site with Pelco Smart Analytics
o
Increasing the tilt angle can help in detecting targets that are directly approaching the
camera.
o
The camera should be tilted no more than 45° from the horizontal.
l
Cameras should be mounted to a stable surface to minimize vibration and movement.
l
Select a lens, mounting height and tilt angle to capture the required level of detail for Classified
Object detection within the scene.
Field of View
l
Camera field of view must be level with the horizon.
l
People in the field of view should be walking upright.
l
People and cars moving parallel to the field of view provide better results than objects moving to or
from the camera.
Object Speed
C6717M-B | 02/23 2
Designing a Site with Pelco Smart Analytics
l
Position cameras so that they can capture moving objects in the field of view for at least 2
seconds.
o
Cameras are designed to detect stationary and moving objects immediately, although there
may be a slight delay in some scenes. However, CCT analytics event configuration requires a
2 second minimum threshold to reduce false alarms.
l
If fast, lateral-moving vehicles are expected, use a wider field of view to increase the available
observation time.
Camera Image Rate
l
For Smart Analytics cameras, there is no minimum image rate for Classified Object video
analytics. The analytics are performed independent of the camera’s encoded image rate setting.
Reflected Light
l
Avoid direct light sources.
o
The camera may be temporarily blinded if bright light sources shine directly at the camera.
l
Position the camera so that the sun, headlights, or other light sources do not shine directly into the
lens.
l
Avoid installing the camera in areas with drastic changes in lighting throughout the day. For
example, avoid installing the camera in an indoor space with direct sunlight through a skylight or
large windows.
o
Significant changes in lighting cause large shadows and different coloring in the space. Such
changes may generate inconsistent detection results.
l
Be conscious of indirect light sources, including reflections from built-in or external IR illuminators,
to avoid lens flares and loss of contrast in the image.
o
Cameras with wide dynamic range (WDR) may be able to overcome this issue in some
instances.
l
Avoid mirrors and other reflective surfaces (like shiny floors and ceilings). Reflections may cause
additional false detections.
Headlights
Headlights can pose a challenge to video analytics, combining low-light conditions with extreme
differences in lighting.
Headlights can interfere with video analytics when:
l
The light shines directly into the camera.
l
The surrounding environment is too dark.
C6717M-B | 02/23 3
Designing a Site with Pelco Smart Analytics
l
The light reflects on wet, snowy, or icy roads.
o
This happens mainly at night, but can also occur during the day when headlights are reflected
into the camera from wet pavement.
l
The light is reflected back at the camera from an enclosed environment, such as a tunnel.
Camera positioning and testing prior to installation are important to minimize reflected light.
l
Position the camera so objects are viewed from the side and not from the front.
l
Add additional illumination (IR or white light) to help balance extreme lighting contrasts.
Contact your Pelco representative for advice on installing cameras when headlights are present in the
field of view.
Adaptive IR
Adaptive infrared (IR) functions by adjusting the IR output dynamically to prevent oversaturation in the
scene as the light changes throughout the night.
l
Cameras using only built-in IR for illumination at night detect targets at a much shorter distance.
Additional illumination is required to consistently detect targets.
l
Be aware that IR may also blur the outline of objects and negatively impact the accuracy of the
video analytics.
o
You can disable adaptive IR to help improve Classified Object detection in the scene.
Lux on Target
l
The recommended minimum illumination is 8 lux on target for analytic cameras.
l
For illuminating distances, it is important to account for lighting, weather, contrast and camera
stability conditions.
o
In bad weather with low visibility, analytics should be combined with other detection methods
to ensure a secure system.
Contact your Pelco representative for advice on installations in challenging lighting situations.
C6717M-B | 02/23 4
Designing a Site with Pelco Smart Analytics
Obstructions
To identify objects accurately, the scene must be clear.
l
For outdoor applications, avoid placing a camera where the field of view includes foliage, terrain or
large objects that occlude the subjects of interest.
o
Also pay attention to obstructions that can reflect infrared (IR) illumination back to the camera
and cause reduced contrast or overexposed video at night. This can be corrected by any of
the following:
l
Separating the IR illuminators.
l
Adjusting the camera placement.
l
Correcting the aim of the IR illuminators or the camera.
l
For indoor applications, a person may be detected as long as their upper body, including head
and shoulders, is visible.
o
It is recommended that a person's full body be visible for the best results.
l
Try to minimize the use of analytics in crowded areas as people are more likely to overlap and
block each other from the field of view. This may cause the system to miss some of the potential
results.
Analytics Scene Mode
In the Camera Configuration Tool, set the camera to use the Analytics Scene Mode that best describes
the scene. Some scene modes are not available on all camera models:
l
Outdoor — suitable for most outdoor environments. This setting optimizes the camera to identify
vehicles and people.
l
Large Indoor Area — only detects people and is optimized to detect people around obstructions,
like chairs and desks, if the head and torso are visible.
l
Indoor Close-up — only detects people and is optimized for scenarios where the camera is
mounted at 4-7 ft (1.2-2.1 m) high and where a person occupies most of the camera’s field of view
and where the full view of the person from head-to-foot is not visible. Examples are ATM
installations or cameras mounted in height strips. This mode does not offer self-learning and does
not support Object Crosses Beam or Direction Violated rules.
C6717M-B | 02/23 5
Designing a Site with Pelco Smart Analytics
Detection Range
l
Install the camera in a location where each object appears in the field of view for at least 2
seconds. Cameras are designed to detect objects within fractions of a second. The exact time
depends on the scene. To be conservative, use 2 seconds or test within your scene.
o
If an analytic rule or alarm uses a region of interest (ROI) or beam crossing to trigger an
event, make sure objects are detected in the camera field of view for at least 2 seconds
before entering the ROI or crossing a beam.
l
If an analytic rule or alarm is based on motion, an object must be moving for at least 1 second to
be classified as moving. The effect is that any object moving less than 1 second is considered
stationary. This initial 1 second is added to any rule threshold time before triggering a motion
based rule.
l
Use the System Design Tool (SDT) to help you estimate the required coverage area for newer
cameras (released in 2022 or later). The SDT is designed to incorporate Pelco video analytic
needs and determines the camera's maximum video analytics detection area in a given scene. To
access the System Design Tool, go to https://sdt.motorolasolutions.com.
For older cameras, use the Pelco Quote Assist Tool to help you estimate the required coverage
area. To access the Quote Assist Tool, go to https://www.pelco.com/quote-assist.
l
Reliable object detection is a function of the camera’s purpose; indoor / outdoor and the prevailing
lighting conditions; daytime / night time / IR assisted. Outdoor analytic scene modes are more
sensitive so will detect a person taking up less of the vertical FOV and therefore will support higher
detection distances. Conversely, indoor analytic scene modes are more resilient to view
obstruction by requiring the person to take up more the vertical FOV for detection.
l
Advanced users can use the following pixel on target recommendations for reliable detection:
Analytics on newer cameras (released in 2022 or later) are designed to operate with a minimum
pixels on target of 9 pixels per foot (30 pixels per meter) based on 2.0 MP resolution. See section
on Camera-Specific Guidelines for guidance about detection requirements based on the height of
a person in the camera field of view.
Your Pelco representative is also a good source for information.
l
Detection range varies by camera and is dependent on the amount of processing available to the
analytics on each camera, and the use case they were designed for. Pelco cameras with less
processing power have a shorter range of detection than those with more processing power.
l
Please be aware that maximum video analytics detection range can change based on the chosen
video aspect ratio, the camera perspective, the focal length and on light and weather conditions.
Further testing in your environment is always recommended.
Outdoor Areas
Be careful not to select a coverage area that is too large, as objects may become obscured by rain or fog
even when there is enough lighting and contrast.
C6717M-B | 02/23 6
Designing a Site with Pelco Smart Analytics
Figure 1: Example of overlapping fields of view for coverage of a building's perimeter.
For perimeter installations:
l
Make sure the camera field of view overlaps to ensure adequate coverage in the blind spot
immediately below a camera.
l
Mount cameras on a central building or structure looking out towards the perimeter.
l
Exceptions:
o
Mount cameras on the perimeter if covering exceptionally large areas.
o
Do not mount on the central building if there is no suitable mounting location, or if there are
obstructions in important areas of the field of view.
C6717M-B | 02/23 7
Designing a Site with Pelco Smart Analytics
Figure 2: Example of overlapping fields of view to provide continuous analytics coverage of an extended
area of interest.
Indoor Areas
Make sure the indoor coverage area is not too small. Low ceilings or confined spaces (such as a security
vestibule between secured doors) may pose problems with establishing a scene that fits the
recommended criteria.
Self-Learning©
Smart Analytic cameras can use Self-Learning algorithms to reduce false detection and alarm rates.
Self-Learning is enabled by default. It allows cameras to actively learn when there is movement in the
scene.
The learning progress requires approximately 200 high-confidence detections throughout the entire field
of view. The time needed to complete the learning progress varies from scene to scene, depending on
the activity in the scene. The algorithm only learns during the day and does not learn if the scene has lowconfidence human activity or low illumination.
C6717M-B | 02/23 8
Designing a Site with Pelco Smart Analytics
In some cases, the Self-Learning Progress Bar may not reach 100%. There may be more false
detections, but true detections will not be affected.
Enable Self-Learning for all video analytics devices, except if:
l
The scene contains objects moving at different heights. For example, an overhead pedestrian
bridge in the background with smaller human activity, compared to larger human activity in the
foreground. Other examples of scenes with objects at different heights are train platforms, mall
stairs and escalators, balconies, hills, and underpasses.
l
The scene contains a significant change in the slope partway in the field of view. For example if a
fence line suddenly goes uphill, it is recommended to disable self-learning or to break up the
scene and position a camera near the bottom of the hill.
Self-Learning can be disabled from the CCT software.
Resetting the Learning Progress
Always reset Self-Learning after a camera is physically moved or adjusted, and if the focus or zoom level
is changed. The change in the camera's field of view affects the video analytic results.
Reset the Self-Learning progress once the camera is stable after initial configuration. During installation,
a camera is frequently adjusted, so any Self-Learning during that time becomes invalid.
You should reset the Self-Learning progress if the position of the home preset is changed on an analytics
capable PTZ camera or, generally, if there are lighting changes.
Self-Learning can be reset from the CCT software.
Camera-Specific Guidelines
Pelco Sarix Professional 4 Series camera
l
Designed for close range analytic detections.
l
For reliable detection, the height of a person should be at least 5% of the height of the vertical
FOV with a maximum of 2/3 height of the vertical FOV.
Pelco Fisheye camera
l
Designed for close range analytic detections.
l
Only people are detected as classified objects. Vehicles and vehicle sub-classes will not be
detected on the fisheye camera.
l
Pelco Fisheye cameras do not offer Self-Learning.
l
Fisheye cameras should be mounted up to a maximum height of 4.5 m (15 ft).
l
To use analytics on the Pelco Fisheye camera, it must be mounted to a flat ceiling parallel to the
floor, and look down on the scene. The camera must be set to Ceiling Orientation Mode.
Analytics are not currently supported on PelcoFisheye cameras that are mounted to a wall or
similar mounting surface and are set to Wall Orientation Mode.
l
The PelcoFisheye camera has an analytics blind spot in the middle 20% of the field of view.
Objects will be detected moving into and out of the blind spot, but will not be detected when they
are inside the blind spot.
Tip: When positioning the camera during installation, try not to install the camera directly above
an area where people often stop to congregate.
C6717M-B | 02/23 9
Designing a Site with Pelco Smart Analytics
l
For reliable detection, the height of a person should be at least 8-15% of the height of the vertical
FOV with a maximum of 2/3 height of the vertical FOV.
Pelco Sarix Corner Mount 3 Series camera
l
Designed for close range analytic detections.
l
Due to the steep angle of the camera lens, there is a detection dead zone directly underneath the
camera for approximately 1.6 ft.
l
For reliable detection for indoor applications, the height of a person should be at least 8-15% of
the height of the vertical FOV with a maximum of 2/3 height of the vertical FOV
Pelco ExSite Enhanced 2 Series camera
l
Designed for long range analytic detections.
l
Analytics are only supported at the home position on the PTZvariant.
l
For reliable detection, the height of a person should be at least 5% of the height of the vertical
FOV with a maximum of 2/3 height of the vertical FOV.
Pelco Esprit Compact Series camera
l
Designed for long range analytic detections.
l
Analytics are only supported at the home position on the PTZ variant.
l
For reliable detection, the height of a person should be at least 5% of the height of the vertical
FOV with a maximum of 2/3 height of the vertical FOV.
l
For reliable PTZ auto-tracking, the height of a person should be at least 5% of the height of the
vertical FOV, approximately 50 pixels.
Pelco Sarix Modular camera
l
Designed for close range analytic detections.
l
Supports Indoor close up analytic scene mode detecting only people as classified objects.
l
Pelco Sarix Modular cameras do not offer Self-Learning.
C6717M-B | 02/23 10
Designing a Site with Pelco Smart Analytics
l
For reliable detection for indoor applications, the height of a person should be at least 8-15% of
the height of the vertical FOV with a maximum of 2/3 height of the vertical FOV.
Pelco Sarix Thermal Enhanced 4 Series camera
l
Designed for long range analytic detections.
l
Use White-Hot or Black-Hot color palette setting for best analytic detection performance. Using
other color palettes may results in additional false detections.
l
Adjust camera image settings for a graduated image contrast. Avoid harsh contrast and reduce
halos around objects for best analytics detection.
l
For reliable detection, the height of a person should be at least 5% of the height of the vertical
FOV with a maximum of 2/3 height of the vertical FOV.
For More Information
If after reading this document you discover that your site requirements deviate from the
recommendations in this document, consult an Pelco representative before installing the cameras. We
may not be able to help you troubleshoot potential issues with Classified Object detection if you do not
follow our recommendations or seek assistance before installing cameras.
To contact a Pelco representative in your area, see: pelco.com/contact-us/.
Pelco Troubleshooting Contact Information
For further assistance, contact Pelco Product Support at 1-800-289-9100 (USA and Canada) or +1-559292-1981 (international).
C6717M-B | 02/23 11
Designing a Site with Pelco Smart Analytics
C6717M-B | 02/23 12
Loading...