Different types of AI, IVS (Intelligent Video Surveillance) & Metadata

 Types of IVS

1. Tripwire Detection

What is Tripwire Detection?

Tripwire detection helps enhance security in surveillance systems by setting up virtual lines on surveillance images. These lines act as tripwires, triggering alarms when someone or something crosses them.

How Does it Work?

1. Setting Up Warning Lines:
   • Users draw warning lines on surveillance images to mark areas of interest.
2. Defining Crossing Direction:
   • Users specify which direction crossing the warning line will trigger an alarm.
   • This direction should match the expected movement of the target being monitored.
3. Alarm Activation:
   • When the target crosses the warning line in the specified direction, the system triggers alarms.
   • Alarms can include sending snapshots and event timestamps to users or activating sirens and lights.

What Can Tripwire Detection Detect?

• Tripwire detection can identify human, vehicle, or both human and vehicle movement.
• It's useful in monitoring narrow passages, entrance/exit points, or roads with clear movement directions.

Considerations for Effective Tripwire Detection:

• For optimal detection, ensure there's enough space on both sides of the warning line.
• Avoid placing warning lines near obstacles or areas with random movement patterns, like plazas.

By following these steps and considerations, users can effectively set up tripwire detection to enhance security and surveillance capabilities.

2. Intrusion Detection (Perimeter Protection)

What is Intrusion Detection?

Intrusion detection enables users to define enclosed polygon areas within surveillance images to monitor specific zones closely. These areas could include sensitive locations or spots requiring heightened vigilance.

How Does it Work?

1. Setting Up Intrusion Detection Zones:
   • Users draw shapes, such as rectangles or irregular polygons, around areas of interest in surveillance images.
   • These defined zones serve as virtual boundaries for monitoring activity.
2. Detecting Intrusions:
   • Intrusion detection can identify two main types of events: when something enters or leaves a defined zone, or when something appears directly within the zone.
   • This feature is commonly used to detect the presence of humans or vehicles.

Practical Applications:

• Intrusion detection is beneficial for monitoring fenced areas, lawns, entrance/exits, and areas where driving is restricted.
• Besides security purposes, users can also utilize intrusion detection for various other applications.

Optimizing Intrusion Detection:

• Ensure the defined boundaries of intrusion detection zones are appropriately positioned within surveillance images.
• Leave some space around the perimeter of the defined zone to avoid missing any activity that may occur near the edges.

3. Loitering Detection

What is Loitering Detection?

Loitering detection helps identify when someone stays in a specific area for an extended time. It's like setting a virtual boundary around an area of interest, and the system alerts you if someone stays there longer than expected. Loitering detection is a valuable feature in surveillance systems, particularly in places like house gates or banks, where people shouldn't linger for long periods.

How Does it Work?

1. Setting Up Loitering Detection Zones:
   • Users define an area of interest by drawing a shape, like a rectangle or irregular polygon, on the surveillance image.
   • This area acts as a virtual boundary where loitering will be monitored.
2. Detecting Loitering Behavior:
   • The system tracks the movement of individuals within the defined area.
   • If someone stays in the area for longer than a set time (minimum alarm time) and their movements don't show significant changes, an alarm is triggered.

Practical Applications:

• Loitering detection is useful for places where prolonged stays are not permitted, like house gates or banks.
• It's primarily used to detect humans lingering in specific areas.

Optimizing Loitering Detection:

• Ensure the defined detection zones cover the entire area of interest.
• Adjust settings, such as the minimum alarm time, to suit the specific surveillance environment and requirements.

4. Fast Moving Detection

What is Fast Moving Detection?

Fast moving detection helps security cameras quickly identify sudden movements, such as someone running in surveillance images. It enables the camera to promptly detect and alert users about potential security incidents.

How Does it Work?

1. Setting Up Fast Moving Detection Zones
   • Users define specific areas of interest by drawing shapes, like rectangles or irregular polygons, on the surveillance image.
   • These areas enable the camera to monitor movement and detect fast-moving objects.
2. Detecting Fast Movements:
   • When someone moves suddenly within the defined area, the camera detects this movement.
   • The system calculates the speed of the moving object and triggers an alarm if it exceeds the user-defined speed limit.

Practical Applications:

• Fast moving detection is useful for identifying individuals attempting to evade surveillance, such as suspects trying to quickly leave the scene.
• It enhances security by allowing for swift detection and response to suspicious activities.

Optimizing Fast Moving Detection:

• Ensure the accuracy of detection by setting the depth of field, which provides essential 3D information for rebuilding the surveillance environment.
• Adjust the speed limit settings based on the specific monitoring scenario and requirements.

5. Illegal Parking Detection

What is Illegal Parking Detection?

Illegal parking detection focuses on identifying instances where motor vehicles are parked in areas where parking is not permitted. It helps enhance security and compliance by promptly alerting users to such violations.

How Does it Work?

1. Setting Up Illegal Parking Zones:
   • Users define prohibited areas by drawing shapes, like rectangles or irregular polygons, on the surveillance screen.
   • These designated zones enable the system to monitor for unauthorized parking activities.
2. Detecting Illegal Parking Events:
   • When a vehicle parks in a prohibited area, the camera captures an on-site snapshot and immediately notifies the user.
   • Optionally, the system can activate a siren and light alarm to warn the vehicle and address the violation.

Practical Applications:

• Illegal parking detection is valuable for enforcing parking regulations in critical areas such as emergency exits, firetruck lanes, and important road entrances/exits.
• It improves safety and security by discouraging unauthorized parking and ensuring compliance with regulations.

Optimizing Illegal Parking Detection:

• Ensure accurate detection by properly configuring the surveillance system's settings, including defining prohibited areas and setting appropriate thresholds for triggering alarms.
• Regularly review and adjust settings based on monitoring needs and changes in the environment.

5. People Gathering Detection

What is People Gathering Detection?

People gathering detection focuses on identifying gatherings or high crowd density in various outdoor and sensitive areas. It utilizes advanced deep learning algorithms to detect human faces and bodies, triggering alarms when abnormal crowd conditions are detected.

How Does it Work?

1. Setting Up People Gathering Zones:
   • Users define areas of interest by drawing shapes, such as rectangles or irregular polygons, on the surveillance image.
   • These designated zones enable the system to monitor for gatherings or high crowd density.
2. Detecting People Gatherings:
   • When a gathering or dense crowd is detected, the system triggers alarms to notify users.
   • This feature helps in scenarios like riots, stampedes, or controlling the spread of epidemics by dispersing crowds immediately upon detection of abnormalities.

Practical Applications:

• People gathering detection is valuable for maintaining crowd control and ensuring public safety in outdoor plazas, government buildings, stations, and other sensitive areas.
• It helps in managing events, traffic flow, and controlling the spread of diseases by promptly identifying crowded situations.

Optimizing People Gathering Detection:

• Configure the surveillance system by defining detection areas, specifying the duration and sensitivity of gatherings that trigger alarms, and adjusting parameters based on specific monitoring needs.

7. Abandoned Object and Missing Object

What is Abandoned Object and Missing Object Detection?

Abandoned object and missing object detection helps identify situations where items are left unattended or taken away without permission. It uses advanced algorithms to analyze surveillance footage and alert users when such events occur.

How Does it Work?

1. Identifying Abandoned and Missing Objects:
   • The system analyzes the video feed and identifies stationary objects in the foreground.
   • It compares these objects with the background to determine if they are abandoned or missing.
2. Triggering Alarms:
   • If an object is detected as abandoned (left unattended for too long) or missing (taken away and not returned), the system triggers an alarm to notify users.

Practical Applications:

• This feature is valuable for railway stations, airports, and other busy areas where items are commonly forgotten or stolen.
• It helps prevent losses and ensures prompt action to retrieve or secure belongings.

Optimizing Detection:

• Configure the system settings to define the duration for triggering alarms and adjust parameters based on the specific monitoring environment.

 Different Types of Metadata

Video metadata refers to descriptive or technical information embedded within or associated with a video file. This metadata provides additional context or details about the video content, facilitating organization, search, and analysis.  As such, all event types utilize Metadata that can be extrapolated in many ways depending on the type of AI feature.

•  Video MetaData: This feature utilizes video Metadata from human and vehicle detected objects. Human metadata, such as clothing type and color can be specified when performing a search, while vehicle traits, such as vehicle type and color can be searched.
• Intelligent Motion Detect 2.0: This advanced form of video detection extrapolates metadata from human and vehicle-recognized objects. This allows the NVR's Object Search function to use the metadata from all iMD 2.0 cameras to display all records of the specified object based on its metadata characteristics.
• Face Detection: Face detection utilizes Metadata acquired from detected faces, including gender, age, and even expression. Search queries can be specified using this metadata.
• Face Recognition: Like Face detection, Face recognition also uses the same data from faces but compares them to added faces in a database for comparison.

Enhancing Security with Siren and Light Alarm Technology

Siren and light alarm technology provides an effective way to deter intruders. By adding a high-decibel speaker and powerful lighting to your camera with our Active Deterrence lineup (ADW cameras), you can transmit warning signals through sound and flashing lights. These components can be combined to create a siren and light alarm system.

Intelligent Video Surveillance (IVS) ensures accurate detection of humans and vehicles, analyzing various intrusion scenarios. This technology allows you to customize alarm sounds and light patterns to deter intruders and prevent incidents effectively.

Additionally, siren and light alarm technology lets you adjust voice formats and alarm rules according to your preferences. You can also fine-tune sound volume to suit different situations, offering versatility in usage.

By configuring alarms for different time periods, you can achieve diverse effects. For instance, you can use welcoming tones during business hours or switch to warning audio during off-hours.

•  See here for more information and setup of Active Deterrence Cameras.