Milesight VS133-P AI ToF People Counting Sensor User Guide

AI ToF People Counting
Sensor
VS133-P
User Guide

BETTER INSIDE, MORE IN SIGHT

Safety Precautions

Milesight will not shoulder responsibility for any loss or damage resulting from not following the instructions of this operating guide.

• Though the device is compliant with Class 1 (IEC/EN 60825-1:2014), please DO NOT look at the ToF sensor too close and directly.
•  The device must not be disassembled or remodeled in any way.
• To avoid risk of fire and electric shock, do keep the product away from rain and moisture before installation.
• Do not place the device where the temperature is below/above the operating range.
•  Do not touch the device directly to avoid the scalds when the device is running.
• The device must never be subjected to shocks or impacts.
• Make sure the device is firmly fixed when installing.
• Do not expose the device to where laser beam equipment is used.
•  Use a soft, dry cloth to clean the lens of the device.

Declaration of Conformity

VS133-P is in conformity with the essential requirements and other relevant provisions of the CE, FCC, and RoHS.

Copyright © 2011-2023 Milesight. All rights reserved.
All information in this guide is protected by copyright law. Whereby, no organization or individual shall copy or reproduce the whole or part of this user guide by any means without written authorization from Xiamen Milesight IoT Co., Ltd.

For assistance, please contact
Milesight technical support:
Email: [email protected]
Support Portal: support.milesight-iot.com
Tel: 86-592-5085280
Fax: 86-592-5023065
Address: Building C09, Software Park
Phase III, Xiamen 361024,
China

Revision History

2. Add multi-device stitching feature;
3. Add installation height detection feature;
4. Add DHCP feature;
5. Display HTTP/MQTT connection status and support data re-transmission feature;
6. Add DST time feature;
7. Add To F frequency setting.

Product Introduction

1.1 Overview
VS133-P is a sensor that uses second-generation ToF technology to accurately count people.
This technology provides more precise depth maps and longer detection distances while maintaining an excellent privacy protection rate. The advanced Tofu technology combined with an AI algorithm enables the sensor to handle complex scenes and distinguish non-human objects with up to 99.8% accuracy. With easy installation, VS133-P is ideal for entrances or corridors in retail stores, malls, offices, subways, and other locations.

1.2 Key Features

• Up to 99.8% accuracy combining the 2 generation ToF technology and AI algorithm
• Allow to collect people counting data by differentiating between children and adults and detecting staffs via identification features for clearer people analysis
• Support the linking of multiple devices to expand coverage
• Wider field angle to obtain longer-distance depth maps and cover a larger area
• Working well even in low-light or completely dark environments with great lighting adaptability
• Free from privacy concerns without image capturing
• Smart U-turn counting to filter redundant counting of people wandering in the area
• High compatibility of data transmission from Ethernet port (HTTP/MQTT/CGI)
• Various serial ports are equipped
• Support local data storage and data retransmission to collect data securely
• Easy configuration via Ethernet port for web GUI configuration
• Quick and easy management with Milesight Device Hub

Hardware Introduction

2.1 Packing List

If any of the above items is missing or damaged, please contact your sales representative.
2.2 Hardware Overview

2.3 Reset Button

2.4 Dimensions (mm)

Power Supply

VS133-P can be powered by 802.3at PoE+. Choose one of the following methods to power up the device.

• Powered by a PoE Switch

• Powered by a PoE Injector

Access the Sensor

VS133-P sensor provides user-friendly web GUI for configuration and users can access it via Ethernet port. The recommended browsers are Chrome and Microsoft Edge. The default IP of Ethernet port is 192.168.5.220 (can be found on the device label).
Step 1: Power on the device and connect the Ethernet port to a PC.
Step 2: Change the IP address of computer to 192.168.5.0 segment as below:
a. Go to Start Control Panel Network and Internet Network and Sharing Center Ethernet Properties Internet Protocol Version 4 (TCP/IPv4).

b. Enter an IP address that in the same segment with sensor ( e.g. 192.168.5.61, but please note that this IP address shall not conflict with the IP address on the existing network);

Step 3: Open the Browser and type 192.168.5.220 to access the web GUI.
Step 4: Select the language.
Step 5: Users need to set the password and three security questions when using the sensor for the first time (three questions can be skipped by refreshing webpage). After configuration, log in with username (admin) and custom password.

Note:

1. Password must be 8 to 16 characters long, which contains at least two kinds or more in combination with numbers, lowercase letters, uppercase letters and special characters .
2. You can click the “forgot password” in login page to reset the password by answering three security questions when you forget the password if you set the security questions in advance.

Operation Guide

5.1 Dashboard
After logging on to the device web GUI successfully, user is allowed to view live video as follows.

Alarm Output: dry contact, output=two contacts closure
Track| When enabled, there is tracking line when people pass the detection area.

Note: When working mode is Node mode, the device will not show people counting data.

5.2 Rule
VS133-P supports 3 working modes:
Standalone Mode: works as a standalone device to count people.
Master Mode: works as a master device to receive live view and tracks from other node devices. One master device can connect 3 node devices at most.
Node Mode: works as a node device to forward live view and tracks to the master device.

5.2.1 Basic Counting Settings
Draw Detection Lines
Users can draw detection lines to record the people count values which indicate the number of people enter or exit.
Step 1: Click Draw Detection Lines.
Step 2: Left-click to start drawing and drag the mouse to draw a line, left- click again to continue drawing a different direction edge, and right-click the mouse to complete the drawing. The line can be dragged to adjust the location and length. One device supports at most 4 broken lines with maximum 4 segments each.
Step 3: If users need to delete the line, click Draw Detection Lines and select the line which need to be deleted, then click Clear This Line or click Clear All.

Note:

1. The arrow direction of the detection line depends on your drawing direction.

2. Ensure that the detected target can pass through the detection line completely. It’s recommended that the detection line is perpendicular to the In/Out direction and on the center of the detection area without other objects around.

3. A redundant identification area needed to be left on both sides of the detection line for the target. This is to ensure that the sensor has stable recognition and tracking of this target before it passes the detection line, which will make the detection and count more accurate.

Deployment Parameters

detect the current installation height automatically.
Note: Dark floor/carpet (black, grey, etc.) does not support to detect the installation height automatically.
Max Target Height| Set the maximum target height, then the device will ignore the objects higher than this setting value.
Min Target Height| Set the minimum target height, then the device will ignore the object shorter than this setting value.
Child Filter Height| Set the max child height when children distinction feature is enabled.

Note: Due to the error in ToF distance measurement (0.035 m), the Max. Target Height should be set as maximum pedestrian height plus 0.035 m and the Min. Target Height as minimal pedestrian height minus 0.035 m in the actual applications. For example, if the pedestrian height is 1.6 m to 1.8 m, the Max. and Min. Target Height should be configured as 1.835 m and 1.565 m respectively.
Counting Strategy

the device will count the In and Out values only when people passed this area. Users can left-click to start the drawing and add edges for this area, then right-click to stop drawing.
Children Distinction| The device will detect the people shorter than child filter height as children.
Staff Detection| The device will detect the people who wear reflective stripes as staff tags on the visible parts (neck, shoulders, etc.) as staffs.
Reflective stripe requirements: width > 2cm, 500 cd/lux.m2
Input Enable Count Externally| Only when alarm input status is the same as the preset trigger status, will the device count the data.
Alarm Input: dry contact, low level=two contacts disconnected, high level=two contacts closure

Report Strategy

value changes.
Periodic Report| Report the people counting data periodically.
Period| Set the period of reporting periodic report.
Range: 1-1080 mins, default: 10 mins

5.2.2 Multi-Device Stitching
Multi-device stitching is mainly used to monitor a larger detection area than just the area covered by a single device. When using this feature, devices should be installed next to each other and ensure the detection areas tangent or overlapping. It only uses one master device to output total counting data.

Before using this feature, set one device as Master Mode and other devices as Node Mode.

Note:

1. Ensure the head of one person can be seen on both live views at the same time.
2. The devices can also be installed without overlapping as required.

Device Positioning
Device positioning is done via X&Y coordinates. For example, the installation direction of the master device is shown as below. When the master device’s coordinate is (0, 0), the coordinates of the node devices are all positive values.

Add Node Devices
Step 1: Go to the master device web GUI, then click Bind Node on Multi-Device List. The device will search for the unbound node devices under the same local network.
Step 2: Select the node device and type the login password of the node device.
Step 3: Fill in the installation height of a node device and relative position information if these parameters are already measured. If not, save default settings and skip to Step 4.

Step 4: Select the node device on the Mult-Device List, click Adjust Relative Position.

Drag the live view of node device to adjust the location and angle, and the relative position parameters will change automatically as your operations. Besides, users can also adjust the size of this live view.

Tips: cut the staff tags or other reflective stripes into pieces and stick them to the ground of overlapping areas, then drag the live view of node devices to make highlight markers in the two live views overlap.
Step 5: Click Set & Testing Track, then check if the tracking lines are connected and smooth when people pass on the live views of multiple devices. If not, click Stop Testing to adjust the node device’s live view location slightly.

Step 6: When all settings are completed, users can draw detection lines and even U-turn areas on the new stitching live view the same as standalone mode devices.
Step 7: Click Unbind to disconnect the node device if necessary.

Node Mode

master device.
Master Device IP Address| Show master device’s IP address. When this IP address is under the same network with node device, the node device can bind to the master device.
Master Device SN| Show the master device’s serial number.
Master Device Name| Show master device name.
Unbind Master Device| Click Unbind to release the connection status, this device will be deleted from the list of the master device.

5.3 Communication
VS133-P provides a Ethernet port for wired access. Besides, users can get the people counting data or configure the device via CGI. For CGI document, please contact Milesight IoT support: [email protected].

192.168.5.220.
Subnet Netmask| Set the Netmask for the Ethernet port.
Default Gateway| Set the gateway for the Ethernet port’s IPv4 address.
Primary DNS Server| Set the primary IPv4 DNS server.
Secondary DNS Server| Set the secondary IPv4 DNS server.
Test| Click to test if the IP is conflicting.

Note: When working mode is the Node mode, the device will not support HTTP/MQTT settings.

Note: When it is being disconnected, the device will store up to 100,000 pieces of data and re-transmit them when the connection is back.
HTTP|
Connection Test| Click Test to send test message to URL to check connectivity.
URL| The device will post the people counting data in json format to this URL.
User| The username used for authentication.
Password| The password used for authentication.
MQTT|
Host| MQTT broker address to receive data.
Port| MQTT broker port to receive data.
Client ID| Client ID is the unique identity of the client to the server.
It must be unique when all clients are connected to the same server, and it is the key to handle messages at QoS 1 and 2.
Username| The username used for connecting to the MQTT broker.
Password| The password used for connecting to the MQTT broker.
Topic| Topic name used for publishing.
QoS| QoS0, QoS1, QoS2 are optional.
TLS| Enable the TLS encryption in MQTT communication.
Certificate Type| CA Signed Server or Self Signed is optional.
CA signed server certificate: verify with the certificate issued by Certificate Authority (CA) that pre-loaded on the device.
Self signed certificates: upload the custom CA certificates, client certificates and secret key for verification.

5.4 Report
VS133-P supports to generate visual line chart or bar chart to display the people traffic and supports to export the report. Before using this feature, ensure that the device time is correct on System page.
Note: When working mode is Node mode, the device will not generate this report.

option.
Export| Export the historical traffic data as CSV file according to the selected time unit. The device can store up to one million data records to CSV file.
Staff
Included/Excluded| Select whether to contain staff counting values on the graph.
| Select the display type as line or bar.
| Download the graph screenshot.

5.5 System
5.5.1 Device Info
All information about the hardware and software can be checked on this page. Besides, users can modify the device name, customize device ID and site ID for large amounts of devices management.

5.5.2 User

the login password of this device.

Security Question| Click to type administrator password, then set three security questions for your device. In case that you forget the password, you can click Forget Password button on login page to reset the password by answering three security questions correctly.

5.5.3 Time Configuration

Start Time: the start time of DST time range.
End Time: the end time of DST time range.
DST Bias: the DST time will be faster according to this bias setting.
Synchronize Mode| NTP Timing or Manual Timing is optional.
Server Address| NTP server address to sync the time.
Time Interval| Set the interval to sync time with NTP server.
Setting Time| Set the device time manually.
Synchronize with computer time| Synchronize the time with your computer.

5.5.4 Remote Management
Users can connect the device to the Milesight Device Hub management platform on this page so as to manage the device centrally and remotely. For more details, please refer to DeviceHubUser Guide. Before connecting, ensure the device has connected to  network via Ethernet port and Internet connection is seamless.

DeviceHub server,
options are Authentication Code and Account.

5.5.5 System Maintenance

Frequency
Adjustment| Adjust the ToF frequency modulation mode to avoid the interference of surrounding IR devices. When using Multi- Device Stitching, please avoid using the same mode with other node devices.
Note: if there is only one option, please contact Milesight IoT support: [email protected]
Reset| Recovery device basic configuration: keep the IP settings and user information when resetting.
Recovery device to factory settings: reset device to factory default, which needs to verify admin password.
Reboot| Restart the device immediately.
Upgrade| Click the folder icon and select the upgrading file, then click the Upgrade button to upgrade. The update will be done  when the system reboots successfully.
Note: The upgrade process takes about 1-10 minutes. Do not turn off the power and complete automatic restart after the  upgrade.
Backup and Restore| Export Config File: Export configuration file.
Import Config File: Click the file icon and select the configuration file, click Import button to import configuration file.

Installation Instruction

Parameter definition:

6.1 Installation Height
The maximum installation height is 3.5 m and the minimum installation height is hmax+d+Δd. For example, when the maximum pedestrian height is 1.8 m, then the minimum installation height is 1.8+0.5+0.035=2.335 m.

6.2 Covered Detection Area
The detection area covered by the device is related to the field of view angle of the device, the installation height and the target height. The length of the detection area is approximately x=2.300×(H-hmin) and the width of the detection area is approximately  y=1.678 x (H-hmin).For example, if the Minimum height of pedestrians is 1.7 m, the detection area corresponding to each installation height is as follows:

6.3 Environment Requirements

• Dark floor/carpet (black, grey, etc.) will affect the device to count staffs when Staff Detection is enabled.

• Avoid 940nm light which may result in incorrect counting.

• Outdoor sunlight shining on the over channel will not have any effect, but the mirrored reflections that allow sunlight to shine on the ToF Sensor should be avoided.

6.4 Installation
Ceiling Mount
Step 1: Ensure the thickness of the ceiling is more than 30 mm, then attach the mounting sticker to the ceiling and drill 4 holes with a diameter of 6mm. If the wire needs to be extended to the interior of the ceiling, a wire hole with a suitable size is also required  to be drilled.
Step 2: Fix the wall plugs into the ceiling holes.
Step 3: Remove the cover on the device, and then connect all required wires and pass them through the wire hole behind the device or block on the side of the device if the wires need to be protruded from the side of the device.
(Note: if the alarm I/O of VS133-P is going to be used, please connect a multi-interface cable to the device)Step 4: Fix the device to the wall plugs via mounting screws; remember to adjust the mounting direction according to the detection area requirement.
Step 5: Fix the cover back to the device.

Ceiling/Lintel Mount (with Optional VB01 Multifunctional Bracket)
Step 1: Attach the mounting plate to the device with 4 screws.
Step 2: Fix the pole to the mounting plate with the hole on the plate.
Step 3: Adjust the length of the pole, then adjust the direction of 3-axis ball and tighten it with the handle.
Step 4: Determine the mounting location and drill 3 holes, fix the wall plugs into the mounting holes, then fix the bracket base to the wall plugs via mounting screws.
(Note: If the wire needs to be extended to the interior of the ceiling or wall, a wire hole with a suitable size is also required to be drilled.)

Step 5: Remove the cover on the device, and then connect all required wires and pass them through the inside of pole.
(Note: if the alarm I/O of VS133-P is going to be used, please connect a multi-interface cable to the device)
Step 6: Fix the pole to bracket base with screws and nuts.

Note:

• Tilt installation should be avoided. Ensure that the front of the device and the ground plane are paralleled.

• Avoid installing the device against the wall and ensure that the device keeps away from the wall at least 30 cm on the short side and 50 cm on the long side.

• Ensure that there are no other objects blocking the ToF light within a 30 cm radius of the front of the device.

• When you install devices on the top of swinging doors, it is suggested to keep the door normally open. If the door must be normally closed, please install the device on the other side of the door to keep away from the door movement. And it is suggested to keep away from the door with a distance of at least 30 cm.

6.5 Factors Affecting Accuracy

• Wearing a fisherman’s hat or carrying a cardboard box on the shoulder: The target will not be recognized because it will become unlike a human in depth map.
• Handheld or cart-carrying a humanoid doll with sufficient height to pass by: The doll will be mistakenly detected as people because it is human-like in depth map.

Communication Protocol

VS133-P will post the people counting data in json format to HTTP URL or MQTT broker.

• Periodic Report

www.ime.de

References

• Support : IoT Support
• IoT & Wireless - ime mobile solutions GmbH +49 6821 91900

Read User Manual Online (PDF format)

Download This Manual (PDF format)

Download this manual  >>