Keep your safety secure with 24-hour monitoring.

TOV-Ai Server

Through intelligent image analysis, it analyzes various events such as object behavior, patterns, and vehicle license plates, etc. and when specific events occur, it automatically displays the event, provides real-time alerts, and enables rapid cause identification.

Intelligent Image Analysis
Intelligent Image Analysis
Event Analysis
Event Analysis
Providing Real time Notification
Providing Real time Notification

TOV-Ai Server Features

Deep Learning Engine
· Providing reliable intelligent analysis through continuous deep learning updates
Minimal False Detection / False Alarms

· Accurate detection even in changing light conditions & diverse environments.

· Smart screening control

· Various event actions (alarm output, image popup, email, HTTP, TCP, FTP, RS485, ONVIF, VMS integration)

Event Relay Server

· Optimized event integration/processing (AI-Server ↔ VMS event transmission/reception & provision of various action functions).

· Various event actions (alarm output, image popup, email, HTTP, TCP, FTP, RS485, ONVIF, VMS integration)

Efficient Resources Management

· When deploying incompatible VMS/AI servers, network & CCTV performance requirements increase, while system stability decreases.

· When providing optimized image relay, network & CCTV performance requirements decrease, while system stability increases.


AI Analysis Type

Object Detection

>
Deep learning-based Image Analysis

>
Intelligent Event Popup

Expected Configuration

AI + VMS(1)AI + VMS(2)

AI Intelligent Image Analysis Solution

[Real time Screen Capture of Deep Learning Intelligent
Image Analysis Solution]
[Intelligent Detection Function Setting Window]

Solutions That Analyze Objects Through Intelligent Image Analysis and Prevent Behavioral & Aafety Accidents

By analyzing image data in real time, it automatically detects and analyzes various behaviors and states of objects within the image, and by storing intelligent events, it provides intelligent event search and event alerts so that it can effectively support achieving diverse objectives, such as enhanced security, optimized safety management, and securing operational data, etc. Based on deep learning technology, it provides the following sophisticated analysis capabilities

Features of Deep Learning Intelligent Image Analysis Solution

· All-in-One System: Integration of video analysis capabilities into AI servers (low cost, high performance)

· Flexible Setting: Activated multiple analysis models per channel & distributed settings of hardware(CPU/GPU/NPU)

· Intelligent Event: Support for displaying and searching events based on META data

· User Convenience: Intuitive UI/UX (zone, threshold setting, etc.)

· Various Alarms: Customized alarm settings based on intelligent events

· Enhanced Safety: Preventing accidents through early detection of hazardous situations and prompt response

· Enhanced operational efficiency: Streamlined workforce management through automated detection and alerts, rapid image search, etc.

· Compliance with Regulations: Strengthening safety management through automated detection of violations such as failure to wear safety helmets, etc.

· Data-based Operation: Decision-making support through objective data(crowding, queues, etc.) 

Features of Deep Learning Intelligent Image Analysis Solution

· All-in-One System: Integration of image analysis capabilities into AI servers (low cost, high performance)

· Flexible Setting: Activated multiple analysis models per channel & distributed settings of hardware(CPU/GPU/NPU)

· Intelligent Event: Support for displaying and searching events based on META data

· User Convenience: Intuitive UI/UX (zone, threshold setting, etc.)

AI Fire Detection(Flame, Smoke) Solution

[Flame/Smoke Image Analysis Solution Real time Screen Capture)
[Flame/Smoke Image Analysis Solution Setting Screen]

Deep Learning Flame/Smoke Intelligent Image Analysis Solution Functions

· Individual settings available of DNN analysis model for each function (flame, smoke) in the reliability and area of interest 

· Event alerts such as image pop-ups, warning sounds, etc. and it can be linked to external equipment such as sirens

· Simultaneous situation updates can be delivered to central control room and multiple distributed control locations. 

· Event notifications can be configured with different conditions for each monitoring location.

Solution to Detect Flame/Smoke Through Intelligent Image Analysis

Analysis of all image input to the NVR enables monitoring across a wide area, allowing users to quickly verify and respond to actual incidents in real time through the image feed.

Specialized Functions of Deep Learning Flame/Smoke Intelligent Analysis Solution

· Equipped with AI deep learning fire detection capabilities, the storage distribution device enables low-cost, high-performance intelligent analysis.

· Through image analysis options such as individual activation of flame/smoke functions by channel, object analysis cycle, object reliability, moving object analysis, and various object analysis models, etc., it has an optimized configuration according to the usage environment and drastically reduces control fatigue due to false alarms.

· Object analysis HW is selectively separated into CPU and GPU to configure HW optimized for the field.

· It provides a user-friendly UI with event duration, object recognition area of interest, and indication of confidence value, etc.

· It has an optimized system configuration supporting various alarms according to fire/smoke events

Deep Learning Fire/Smoke Event Search Support

It supports event search through AI intelligent image analysis data


AI License Plate Recognition(LPR) Solution

[Intelligent Deep Learning LPR Solution Real time Screen Capture]
[LPR Function Setting Window]

Deep Learning Intelligent Image Analysis Solution Activation

· It analyzes vehicles by distinguishing forward/reverse direction based on their travel direction.

· It can search by date/time/camera/operation information/license plate (full or partial).

· Event actions such as image popup and alarms can be triggered based on the configured license plate.

· Simple and intuitive setup and search methods

It analyzes images input to the NVR in real time for LPR.

It simultaneously performs analysis, license plate storage, image storage, and information display on screen, and it is possible to search alongside image footage, which makes it adaptable for diverse applications across various sites such as roads, urban security, and apartment complexes, etc.

Specialized Functions of Deep Learning LPR Solution

· All-in-one low-cost, high-performance system featuring LPR function on NVR

· Individual settings for LPR by channel

· It analyzes vehicle directionality using intelligent analysis to detect wrong-way driving events.

· Object analysis hardware separated into CPU & GPU for an on-site optimized hardware configuration

· It provides a user-friendly UI, such as displaying license plate data, LPR locations, and text display, etc.

· System configuration according to purpose of use for specific events (whitelist, blacklist)

Deep Learning Fire/Smake Event Seach Support

It supports event search through AI intelligent image analysis data.


Server Specifications

(Individual Specifications)

Product Features

· Deep learning based image analysis server

· Image analysis available while maintaining TOVNET NVR/VMS

· Alarm monitoring system integration interworked with VMS, NVR

· Support for various APIs and over 130 protocols

· Support for Max. 8K monitor output

Model Name
EM-AI04mini
EM-AI08mini
EM-AI16mini
EM-AI00mini
EM-AI16S
EM-AI32S
EM-AI00S
Image Input
Default
4CH
8CH
16CH
0CH
16CH
32CH
0CH
Max. License
up to 4CH
up to 8CH
up to 16CH
up to 32CH
CPU
intel i5(14cores)
dual intel i5(14+14 cores)
GPU
intel Arc Graphics
dual intel Arc Graphics
NPU
intel NPU3(11TOPs)
dual intel NPU3(22TOPs)
Total TOPs(CPU+GPU+NPU)
up to 34TOPs
up to 68TOPs


Inferential Limit
20202424484848
RAM
DDR5 16GB
DDR5 32GB
DDR5 64GB
SSD
M.2 SSD 128GB(OS, Event DB)
dual M.2 SSD 128GB
Display
Monitor Output
1 x HDMI(8K60Hz) + 1 x HDMI(4K60Hz)
Split Screen
(1/2/4)
(1/2/4/6/9)
(1/2/4/6/10/13/16)
depend on CH
(1/2/4/6/10/13/16)
(1/2/4/6/10/13/16)
depend on CH
Performance
Live
FHD 4CH
FHD 8CH
FHD 16CHup to license
FHD 32CH
up to license
Storage
Event Data Only
Network
Interface
2.5GbE 1port, 5GbE 1port
Physical Bandwidth
7.5Gbps
Max. Users
1024(within network bandwidth)
Protocol
TCP/IP,UDP,DNS,DDNS,HTTP,NTP,RTP/RTCP,RTSP
Bandwidth Control
Automatic, Manual

Application Method of Intelligent Screening Control Solutions

Optimized Streaming Sharing + Event Representation Between TOVNET Products

Channel 1
Channel 2
Channel 128

Easy to Set Up Flexible Intelligent Analytics for Each Channel

Object, Behavior, PPE
Fire/Smoke
License Plate Recognition

Easy to Set Up Flexible Intelligent Analytics for Each Channel