SmartVision is a video surveillance and video analytics software platform designed for long-term operation, scalability, and real-world use rather than demo setups. The system was engineered as a tool for engineers and operators, with a primary focus on stability, predictability, resource control, and reducing manual operations.
Connecting and Working with IP Cameras
The SmartVision video surveillance system supports IP cameras streaming video via RTSP or HTTP. It works reliably with both professional-grade and budget cameras, including devices with unstable firmware or network quirks.
Automatic discovery of cameras on the local network is performed via ONVIF. Cameras are detected without manual entry of addresses, profiles, or stream paths, which reduces deployment errors and speeds up system launch. After a camera or network reboot, SmartVision automatically restores the connection.
For cameras using DHCP, an automatic IP address update mechanism based on MAC address is implemented. This solves the common issue of “lost” cameras after equipment restarts. In case of temporary network outages or unstable connections, the system recovers automatically without operator intervention.
Mass Configuration and Camera Management
In SmartVision, a camera is treated not as a simple video stream but as a full-fledged system object with all related parameters. Cameras can be duplicated entirely, including:
This allows you to configure a reference camera once and then apply its configuration to dozens of identical devices. This approach reduces errors caused by repetitive manual actions and simplifies system scaling.
Visualization and Interface
The camera grid is the operator’s primary tool. SmartVision uses updated layout logic where the video stream automatically scales to the tile size. The image is not cropped, does not jump, and looks consistent across different monitors.
Any camera can be expanded to full screen with a double-click. The interface requires no training and behaves predictably, without hidden modes or non-standard actions.
Archive and Data Storage
SmartVision supports archive recording to multiple local disks, network storage devices, and NAS. For each storage target, you can define:
This prevents typical failures related to system disk overflow and allows building a reliable data storage scheme.
Video is recorded directly in the open MP4 format using H.264 and H.265 codecs. The archive is not tied to a proprietary file system or a special player. Disks can be connected to any computer and recordings can be opened with standard media players.
Files are structured by camera and date. Stored separately are:
Recording Modes and Performance
SmartVision offers two approaches to video recording:
In native mode, CPU load is reduced and GPU acceleration is used, allowing the system to handle a large number of cameras without overload. The built-in player ensures correct playback of H.265 recordings.
The choice of mode depends on the task: compatibility or performance.
Continuous Recording and Timelapse
The system supports:
Timelapse allows preserving the history of a site with minimal disk space usage. When activity is detected, the system automatically switches to normal recording and returns to timelapse after the event ends.
This mode is used at construction sites, industrial facilities, remote locations, and in urban scenarios.
Event Recording and Motion Detection
SmartVision records events based on intelligent motion detection. The algorithms are optimized to react to real scene changes rather than shadows, glare, or noise.
This reduces false alarms and makes events suitable for practical use rather than “just in case” viewing.
Video Analytics and Object Recognition
The system performs detection and classification of objects in the frame, including:
Analytics can be used both for security and for accounting, monitoring, and activity analysis.
License Plate Recognition
SmartVision automatically recognizes vehicle license plates, logs events, and generates reports, including parking duration tracking.
Different neural network models are used for different regions. The system automatically determines the country of the plate and selects the optimal algorithm. Specific features of Eastern European license plates are taken into account.
Recognition parameters are configured via configuration files, allowing you to balance speed and accuracy for specific shooting conditions.
Working with Faces and People
SmartVision provides:
The system supports presence reports, recording entry and exit times and forming a clear timeline of a person’s presence on site. These data are used for attendance analysis, working time tracking, and access control.
Speech Recognition (ASR)
Optionally, SmartVision supports continuous audio recording from IP cameras with automatic speech recognition in more than 100 languages. Audio is converted into searchable and analyzable text.
Smoke and Fire Detection
The system performs visual detection of smoke and fire based on scene behavior analysis. This provides an additional layer of protection and can identify signs of fire earlier than physical sensors.
The module is resistant to interference such as fog, steam, and glare, and reliably resumes operation after camera reboots or network failures.
PTZ Control
SmartVision supports PTZ camera control. Pan, tilt, and zoom commands are processed stably and predictably. Latency when working with presets is minimized, which is important for оператив control of the viewing area.
Multi-Server Architecture
Behind its compact interface, SmartVision hides a multi-server architecture designed for enterprise and distributed installations. The system can be used in large-scale deployments and data centers, with load distribution and centralized management.
Remote Access
SmartVision integrates with cloud video surveillance services (VSaaS) and provides remote access to video archives and live streams.
Automatic discovery of cameras on the local network is performed via ONVIF. Cameras are detected without manual entry of addresses, profiles, or stream paths, which reduces deployment errors and speeds up system launch. After a camera or network reboot, SmartVision automatically restores the connection.
For cameras using DHCP, an automatic IP address update mechanism based on MAC address is implemented. This solves the common issue of “lost” cameras after equipment restarts. In case of temporary network outages or unstable connections, the system recovers automatically without operator intervention.
Mass Configuration and Camera Management
In SmartVision, a camera is treated not as a simple video stream but as a full-fledged system object with all related parameters. Cameras can be duplicated entirely, including:
- video streams and codecs,
- detectors and zones,
- rules and events,
- recording schedules,
- analytics parameters,
- cloud and local scenarios.
This allows you to configure a reference camera once and then apply its configuration to dozens of identical devices. This approach reduces errors caused by repetitive manual actions and simplifies system scaling.
Visualization and Interface
The camera grid is the operator’s primary tool. SmartVision uses updated layout logic where the video stream automatically scales to the tile size. The image is not cropped, does not jump, and looks consistent across different monitors.
Any camera can be expanded to full screen with a double-click. The interface requires no training and behaves predictably, without hidden modes or non-standard actions.
Archive and Data Storage
SmartVision supports archive recording to multiple local disks, network storage devices, and NAS. For each storage target, you can define:
- storage space limits,
- recording priority,
- switching scenarios when the storage is full.
This prevents typical failures related to system disk overflow and allows building a reliable data storage scheme.
Video is recorded directly in the open MP4 format using H.264 and H.265 codecs. The archive is not tied to a proprietary file system or a special player. Disks can be connected to any computer and recordings can be opened with standard media players.
Files are structured by camera and date. Stored separately are:
- continuous recordings,
- event recordings,
- timelapse recordings.
Recording Modes and Performance
SmartVision offers two approaches to video recording:
- a universal mode with transcoding to H.264 for maximum compatibility,
- a native mode that records the stream “as is,” including H.265.
In native mode, CPU load is reduced and GPU acceleration is used, allowing the system to handle a large number of cameras without overload. The built-in player ensures correct playback of H.265 recordings.
The choice of mode depends on the task: compatibility or performance.
Continuous Recording and Timelapse
The system supports:
- continuous video recording,
- continuous timelapse recording with configurable frame rate.
Timelapse allows preserving the history of a site with minimal disk space usage. When activity is detected, the system automatically switches to normal recording and returns to timelapse after the event ends.
This mode is used at construction sites, industrial facilities, remote locations, and in urban scenarios.
Event Recording and Motion Detection
SmartVision records events based on intelligent motion detection. The algorithms are optimized to react to real scene changes rather than shadows, glare, or noise.
This reduces false alarms and makes events suitable for practical use rather than “just in case” viewing.
Video Analytics and Object Recognition
The system performs detection and classification of objects in the frame, including:
- people,
- vehicles,
- animals,
- birds,
- other object types.
Analytics can be used both for security and for accounting, monitoring, and activity analysis.
License Plate Recognition
SmartVision automatically recognizes vehicle license plates, logs events, and generates reports, including parking duration tracking.
Different neural network models are used for different regions. The system automatically determines the country of the plate and selects the optimal algorithm. Specific features of Eastern European license plates are taken into account.
Recognition parameters are configured via configuration files, allowing you to balance speed and accuracy for specific shooting conditions.
Working with Faces and People
SmartVision provides:
- face detection,
- identification of predefined persons,
- automatic creation of a database of all detected faces,
- search for similar people across different cameras without personal identification.
The system supports presence reports, recording entry and exit times and forming a clear timeline of a person’s presence on site. These data are used for attendance analysis, working time tracking, and access control.
Speech Recognition (ASR)
Optionally, SmartVision supports continuous audio recording from IP cameras with automatic speech recognition in more than 100 languages. Audio is converted into searchable and analyzable text.
Smoke and Fire Detection
The system performs visual detection of smoke and fire based on scene behavior analysis. This provides an additional layer of protection and can identify signs of fire earlier than physical sensors.
The module is resistant to interference such as fog, steam, and glare, and reliably resumes operation after camera reboots or network failures.
PTZ Control
SmartVision supports PTZ camera control. Pan, tilt, and zoom commands are processed stably and predictably. Latency when working with presets is minimized, which is important for оператив control of the viewing area.
Multi-Server Architecture
Behind its compact interface, SmartVision hides a multi-server architecture designed for enterprise and distributed installations. The system can be used in large-scale deployments and data centers, with load distribution and centralized management.
Remote Access
SmartVision integrates with cloud video surveillance services (VSaaS) and provides remote access to video archives and live streams.
SmartVision initially builds a local video archive on a standard PC or server. The system does not require constant streaming of video to the internet and continues to operate fully even with unstable connections. The cloud is used only where it makes sense: for remote access, backup storage of events, and viewing important fragments. This removes dependency on the communication channel and protects the archive from losses during temporary internet outages.
Cloud Video Surveillance Service
SmartVision supports a flexible cloud video surveillance model in which the local system and the cloud complement rather than compete with each other. This approach maintains data control, reduces costs, and still provides convenient remote access and backup of key events.
Local Video Analytics Without Data Leakage
All video analytics in SmartVision are performed locally. Object detection, face and license plate recognition, motion detection, smoke and fire detection all happen on the user’s side. No unnecessary data is sent to the cloud, and sensitive information remains fully under the system owner’s control.
This approach increases security and simplifies compliance with privacy requirements. Video containing people is not stored in third-party data centers unless necessary, and decisions about what is sent to the cloud are made by the system itself based on events.
Efficient Storage Usage with AI
SmartVision uses computer vision and machine learning algorithms to reduce the volume of stored data. The system can:
This is especially important for cloud storage, where every gigabyte has a cost. Video data is transmitted over secure channels, without opening internet ports and without direct access to the local network.
Problems of Classic Cloud Systems and an Alternative Approach
Traditional cloud video surveillance services usually rely on continuous transmission of the entire video stream to the cloud. In practice, this leads to several issues:
SmartVision uses a different model. The local system continues recording regardless of internet status. Only events are sent to the cloud, not the continuous stream. This reduces bandwidth usage, lowers storage volume, and preserves only fragments with real practical value.
Event-Based Cloud Recording
In the premium version of SmartVision, integration with cloud video surveillance is available. Cloud recording is triggered by events: object detection, face and license plate recognition, smoke, sound, zone intrusion, and other triggers.
As a result, the cloud archive consists of short, informative clips rather than endless hours of background footage. This simplifies viewing, searching, and analysis, while also reducing storage costs.
Online Viewing of IP Cameras
SmartVision supports live viewing of connected IP cameras via the cloud. Users gain access to real-time situational awareness without the need to store weeks of empty recordings in the cloud.
This format is suitable for both private use and business: there is always a way to quickly assess the situation on site without overpaying for a meaningless archive.
Hybrid Model
The SmartVision cloud service is built on a hybrid model: local AI ensures speed, stability, and privacy, while the cloud provides accessibility, scalability, and event backup. This allows you to benefit from the cloud without its typical drawbacks and maintain full control over the video surveillance system.
All video analytics in SmartVision are performed locally. Object detection, face and license plate recognition, motion detection, smoke and fire detection all happen on the user’s side. No unnecessary data is sent to the cloud, and sensitive information remains fully under the system owner’s control.
This approach increases security and simplifies compliance with privacy requirements. Video containing people is not stored in third-party data centers unless necessary, and decisions about what is sent to the cloud are made by the system itself based on events.
Efficient Storage Usage with AI
SmartVision uses computer vision and machine learning algorithms to reduce the volume of stored data. The system can:
- reduce frame rate when there is no motion,
- filter out “empty” scenes,
- store only meaningful fragments instead of hours of repetitive video.
This is especially important for cloud storage, where every gigabyte has a cost. Video data is transmitted over secure channels, without opening internet ports and without direct access to the local network.
Problems of Classic Cloud Systems and an Alternative Approach
Traditional cloud video surveillance services usually rely on continuous transmission of the entire video stream to the cloud. In practice, this leads to several issues:
- data loss during internet outages,
- archive overflow with background video without events,
- a large number of false notifications,
- high costs when scaling,
- privacy and personal data storage concerns.
SmartVision uses a different model. The local system continues recording regardless of internet status. Only events are sent to the cloud, not the continuous stream. This reduces bandwidth usage, lowers storage volume, and preserves only fragments with real practical value.
Event-Based Cloud Recording
In the premium version of SmartVision, integration with cloud video surveillance is available. Cloud recording is triggered by events: object detection, face and license plate recognition, smoke, sound, zone intrusion, and other triggers.
As a result, the cloud archive consists of short, informative clips rather than endless hours of background footage. This simplifies viewing, searching, and analysis, while also reducing storage costs.
Online Viewing of IP Cameras
SmartVision supports live viewing of connected IP cameras via the cloud. Users gain access to real-time situational awareness without the need to store weeks of empty recordings in the cloud.
This format is suitable for both private use and business: there is always a way to quickly assess the situation on site without overpaying for a meaningless archive.
Hybrid Model
The SmartVision cloud service is built on a hybrid model: local AI ensures speed, stability, and privacy, while the cloud provides accessibility, scalability, and event backup. This allows you to benefit from the cloud without its typical drawbacks and maintain full control over the video surveillance system.