Control Panel: How to add cameras to Vision and enable AI services for the added cameras
Initial steps
To add a camera, follow these steps (Refer to Figure 1):
- Navigate to the Control Panel
, located in the bottom right corner of the starting interface.
- Open the Control Panel menu and select the Cameras tab.
- In the Cameras tab, click the "+" sign
to open three options.
- Options are:
Add Cameras Manually
Add Cameras Automatically
Add USB Cameras Manually
Figure 1 – Cameras tab
Option 1: Adding cameras automatically
Select the location from a drop-down menu and click the "Scan" button (refer to Figure 2).
Figure 2 – Adding cameras automatically
After the initial scan, all cameras on your network should be visible. Select the cameras you wish to add to Vision and click the "Add" button (refer to Figure 3).
Figure 3 – Scan results
Once added, the camera will appear in the Cameras tab. Click on the newly added camera to access its details (refer to Figure 4).
Figure 4 – Accessing camera details
In "Camera Details", the required fields in the "Network" section will be automatically filled. Click the "Connect" button (refer to Figure 5).
Figure 5 – Camera Details
After clicking the "Connect" button, you will be presented with the Authentication window. Here, enter the credentials of the camera you are trying to add and click the "Authenticate" button (refer to Figure 6).
Figure 6 – Authenticate
You have successfully added a camera to Vision. If everything is alright, you will see the camera stream. If not, please check the "Log" tab to find out what's causing the issue and debug.
Option 2: Adding cameras manually
When adding a camera manually, choose the location for which you’re adding the camera. Enter the number of cameras you wish to add and press the "Add" button (refer to Figure 2).
Figure 2 – Add cameras manually
The cameras will be added and visible on the Cameras tab. Find the camera you added and click it to enter the camera details (refer to Figure 3).
Figure 3 – Camera Details
In "Camera Details", the required field in the "Network" section is the IP address. Enter the IP address of the camera and click the "Connect" button (refer to Figure 4).
HTTP and RTSP ports will be automatically set to 80 and 554, but you can change them if necessary.
Figure 4 – Entering camera details manually
After clicking the "Connect" button, you will be presented with the Authentication window. Here, enter the credentials of the camera you are trying to add and click the "Authenticate" button (Refer to Figure 5).
Figure 5 – Authentication
You have successfully added a camera to Vision. If everything is alright, you will see the camera stream. If not, please check the "Log" tab to find out what's causing the issue and debug.
RTSP URLs
Note: Applicable to options 1 and 2 only.
Adding RTSP URLs manually is typically optional, as the system automatically completes this process in most cases.
The process of adding RTSP URLs manually varies based on specific scenarios:
- If the system has automatically generated the RTSP URLs, manual input is unnecessary.
- In cases where the system has auto-filled the RTSP URL fields but a user prefers to use a different RTSP URL (due to the availability of multiple URLs for some cameras), manually editing the URL is optional.
- Manual addition becomes necessary only if the system fails to generate the RTSP URL, which is a rare scenario.
After authentication, the RTSP URL fields in the "Camera details" will be automatically populated based on the camera details. The system should automatically generate primary and secondary RTSP URLs, with the primary as default for previewing streams in best quality.
However, RTSP URLs can also be manually edited. - this is an optional step and may be necessary under certain circumstances for specific reasons.
Steps to manually create and add RTSP URL
In some cases, you may for X reason prefer to manually create and enter the RTSP URL instead of having the system automatically generate it. In such cases, follow these steps:
How to create an RTSP URL?
An example of an RTSP URL for HIKVISION cameras:
rtsp://USERNAME:PASSWORD@IPADDRESS:RTSPPORT/Streaming/Channels/101 <- Primary RTSP URL
rtsp://USERNAME:PASSWORD@IPADDRESS:RTSPPORT/Streaming/Channels/102 <- Secondary RTSP URL
Or find an RTSP URLs for your camera model here.
For most cameras, the RTSP port will always be 554.
Option 3: Adding USB camera manually
We've recently added the option to add a USB camera manually. This process is similar to adding other cameras manually. You'll need to provide the necessary details and click the "Connect" button.
Deleting cameras from Vision
To delete a camera from Vision, click on the “Delete” button and then confirm the delete (refer to Figure 7).
Figure 7 – Delete camera
AI services in Vision system
AI Services within the Vision system are tools that apply artificial intelligence for real-time video feed analysis. These include features like person counting, face recognition, and age and gender prediction. Additionally, to ensure privacy, the Vision system complies with GDPR regulations using features such as Identity Anonymize.
Understanding AI Services settings
AI services refer to advanced features that use artificial intelligence to analyze video feeds in real time. Here's a breakdown of the AI services:
- AI Processing: The application of artificial intelligence in analyzing video feeds. When AI processing is enabled, the system uses machine learning algorithms to interpret the video data it receives.
- People Counting: This feature uses AI to count the number of people in the camera's field of view. It can be used in various settings, such as retail stores to track customer traffic, offices to monitor occupancy, or public spaces to manage crowd sizes.
- Face Recognition: This AI service can identify or verify a person by comparing and analyzing patterns based on the person's facial contours. It is commonly used for access control in buildings, or for identifying individuals in a crowd.
- Age and Gender Prediction: These features use AI to estimate a person's age and gender based on their facial features. This can be useful in demographic studies or targeted marketing campaigns.
- Identity Anonymize: In compliance with privacy regulations like GDPR, this feature obscures the identities of individuals captured in the video feed. This is especially important in public spaces where privacy is a concern.
Some of these AI services require substantial computational resources, and the accuracy of their results can depend on various factors, including the quality of the video feed and the specific algorithms used.
Enabling AI services
To enable the AI service, navigate to the AI Services tab in the camera details and click Enable AI Processing (refer to Figure 8):
Figure 8 – AI services
When AI processing is enabled, People Counting will always be active, while face recognition, age, and gender prediction can be turned on or off.
Anonymization of identity in live view
We've gone a step further in making the Vision solution fully GDPR compliant by adding an anonymization feature. The identity anonymization feature allows you to exclude the personal information of recognized identities in Live View . However, we recommend that you select this option only if required by the regulations in your country.
Figure 9 – Identity Anonymize feature
Understanding identity Creator and Validator
If you set the "Camera Identity" option, in the details of a certain camera, as Identity Validator, it means that this camera will not create new identities but only verify individuals by comparing them with existing identities in the system (refer to Figure 10).
Figure 10 – Camera Identity Type
Identity Creator camera
The Identity Creator settings in the details of a certain camera refers to the camera's ability to create a new identity profile based on an individual's facial features. Additionally, the sliders associated with the Identity Creator adjust the following parameters:
- Minimum face size: This is the smallest face size (in pixels) that the system will consider when creating a new identity. Faces smaller than this size will be ignored.
- Minimum recognition (face exposure) time: This is the minimum amount of time (in milliseconds) that the system needs to observe a face before it can create a new identity. This ensures that the system has enough data to create a reliable profile.
- Minimum face orientation angle: This is the minimum angle at which the system can reliably recognize a face. Faces that are turned away from the camera beyond this angle are not recognized reliably.
Figure 11 – "Identity Creator" Adjustment
Identity Validator camera
The Identity Validator setting in the details of a specific camera enables the verification of an individual by comparing them with existing identities in the system, however, a camera configured as a Validator will not be able to create new identities. Additionally, the sliders associated with the Identity Validator adjust the following parameters:
- Minimum face size: Similar to the Identity Creator, this is the smallest face size that the system will consider when matching a face to an existing identity.
- Minimum recognition (face exposure) time: This is the minimum amount of time that the system needs to observe a face before it can match it to an existing identity.
- Minimum face orientation angle: This is the minimum angle at which the system can reliably match a face to an existing identity.
Figure 12 – "Identity Validator" Adjustment
These parameters ensure that the system can create and validate identities accurately and reliably.
Mapping cameras in Vision system
The "Camera Mapping" option within the camera details of a selected camera is crucial for the "Time clock" feature to function. This is where you define the role for that camera in your system. Three options (1. Default, 2. Entry Camera, 3. Exit Camera) correspond to the specific function each camera performs in monitoring movement in and out of the location. Here's a more detailed explanation:
- Default: When you initially create a camera in the system, it is automatically set to the 'Default' configuration (role). This setting implies a flexible use, rather than being specialized for tasks such as monitoring entrances or exits to gather Timeclock data on identities.
- Entrance Camera: This camera is positioned to monitor an entrance. It primarily captures individuals entering a location.
- Exit Camera: This camera is positioned to monitor an exit. It primarily captures individuals leaving a location.
Figure 13 – Camera Mapping
Implementing "Camera mapping" for "Time clock" functionality
The "Camera mapping" options within the Camera Details window will be grayed out if the Time clock feature is not enabled at the space level. To enable Timeclock:
- Follow the steps outlined in here.
- If you are not a Space Owner, contact your space's owner to enable the Time clock feature.
Once Time clock is enabled on your space, you can set up your camera's "mapping" to accurately track individual movements, which is essential for the functionality of the Time clock.
Camera Mapping roles
For example, when a recognized individual passes an "Entrance Camera", it logs a "Check-In" time. Similarly, when they pass an "Exit Camera", it logs a "Check-Out" time. The system then uses these times to calculate the "Total Active Period" (the time spent inside the location) and the "Total Off Period" (the time spent outside the location).
Time clock data
If none of these options are enabled on any camera, the system will not be able to track the movement of individuals in this way, which is why the "Time clock" section would display "Currently no data" in the identity details on the Directory page. The ability to filter by a camera in the directory allows you to review movement data for specific cameras, which can be useful for analyzing traffic patterns at different entrances/exits. Check out the Directory Guide to learn how to interpret your identities' Time clock Data.