Software

For video management, KL-VMS+ (VMS Plus) is KDT's licensed VMS for large projects with several servers; for analytics, KL-PRS reads license plates and KL-FRS recognizes faces. For everyday work with KDT NVRs and cameras, the product specification tables list the KDT CMS/VMS Windows software and apps for Android and iOS. The download center of this website adds device-search tools, KDT Player and KDT Desk, the remote-support program.

A VMS (video management software) is installed on a PC or server and manages many cameras and NVRs from one place, with separate users and permissions, alarm rules, maps, several monitors and analytics. The NVR's web page is for configuring and viewing that single device, and the mobile app is for quick live view and playback. Supported browsers depend on the model: the specification tables list Chrome, Edge, Firefox and IE for most models and IE only for some.

CMS (central management software) is usually the client program that comes with recorders: it connects to several devices and provides live view, playback and export. A VMS is a fuller platform that adds recording servers, multi-server and multi-site operation, user roles, alarm rules and analytics. Vendors do not always keep the two names apart, which is why the specification tables of KDT NVRs simply list 'KDT CMS/VMS'; for the fuller platform, see the KL-VMS+ datasheet.

For one site with a single NVR and a few viewers, the web page and the mobile app are enough and nothing has to be installed. For several recorders or a control room with several monitors, use the Windows KDT VMS; its manual states that it can show 128 cameras across several monitors. For projects with several sites, user roles, maps, alarm rules, or integrated plate and face recognition, choose KL-VMS+, which according to its datasheet has no limit on the number of servers or users.

KL-VMS+ is a client-server VMS for Windows that supports multi-server installation, Persian and English interfaces, Solar (Jalali) and Gregorian calendars, sign-in through Active Directory and user groups with separate permissions. It records on schedule or on motion, plays back on a timeline with thumbnails at up to 127x speed, retrieves footage from camera memory (Edge Recording) and exports single or grouped cameras in standard formats. Its datasheet also lists maps, an alarm rule engine, PTZ presets and patrols, privacy masks, multi-monitor layouts and group configuration of cameras.

According to the datasheet, the KL-VMS+ license is perpetual, places no limit on the number of servers or users, and leaves the camera count to the hardware; version updates are free during the support period. Optional modules are licensed separately: ONVIF support (KL-VMS/O+), integrated license plate recognition (KL-VMS/P+), integrated face recognition (KL-VMS/F+), heatmaps and video analytics such as object detection and classification and virtual fence. The license can be activated online or offline, so a server without internet access can still be licensed.

According to the datasheets, the KL-VMS+ server and client and also KL-FRS run on Windows, and KL-PRS runs on Windows x64 or Linux ARM; all three list virtual machine support. A virtual machine works well for the server role when it gets dedicated CPU cores and memory (no allocation beyond the physical capacity), recordings go to separate disks or a storage network, and clients that decode many streams usually run on physical PCs with a graphics card. KDT VMS is a 64-bit Windows program, and a macOS client package for it is also listed in the download center.

According to the KL-VMS+ datasheet, the connection to the main server and the server-to-client traffic are encrypted, storage and archives can be encrypted, and cameras are reached over secure HTTPS/TLS channels; KL-PRS and KL-FRS also support RTSPS for camera streams. Users sign in through Active Directory, local Windows accounts or the software's own accounts, and each user's activity is logged. The datasheets state that KL-VMS+ and KL-PRS hold an AFTA security certificate and that the AFTA approval of KL-FRS is under review; certificate status can change, so check the latest datasheet for tenders.

KL-PRS is client-server plate recognition software for Windows x64 or Linux ARM that works with cameras of any brand (more features with KDT cameras). It stores each recognition with a snapshot and a 7 to 10 second clip from before and after it, offers searching and playback with many filters, and exports reports to PDF or Excel. It supports national, motorcycle, free-zone and Omani plates, blocking travel on days that do not match the plate's even or odd number, restrictions by province or plate type, skipping parked cars, notes on each record, and, in newer versions, switching between parking and highway modes without a new license. The number of defined users is unlimited, with up to 32 connected at the same time.

KL-FRS is client-server face recognition software for Windows. Faces are enrolled for each person, and people can be allowed or blocked individually, by group or by camera channel; the software can trigger network or USB relays, send SMS messages, show unknown and enrolled people separately in live view and estimate gender, expression and age group. Detection zones, time windows, the required face size (in pixels per meter) and the accepted face angle can be adjusted. Like KL-PRS, it stores each detection with a snapshot and a 7 to 10 second clip, exports PDF or Excel reports and accepts up to 32 simultaneous users.

According to the KDT VMS PRO manual (64-bit Windows), the package has four parts. KDT Client is for display and settings, KDT Recorder sets up recording and allocates part of the PC's storage to recording files, KDT Event handles event settings, and KDT Server shows the server's IP address so that clients can connect to the main server. On a single PC the client connects to that same PC, and in client-server mode it connects to the IP address of the main server.

In the software's Control Panel open Device and click Add. Search finds the cameras and devices on the network so you can select and add them; to add one manually, set the type to IPC (camera) or NVR and enter a name, the IP address and the device's user name and password. Search normally finds only devices in the same subnet, so add devices on other subnets by IP address. KL-VMS+ can also add cameras singly, in groups or by an IP address range.

In Control Panel, Preview, use Add Screen once for each monitor, name each screen and build its layout; double-clicking its name in the top bar detaches it from the software, and you drag it to the target monitor. Monitor Plan builds a rotating display, for example a single-view layout that switches between four cameras on the main stream every 10 seconds, and can run different plans at set times of the day. After building a plan, click the Monitor Plan icon at the bottom of the Preview page to run it.

According to the KDT VMS manual, first format one or more drives of suitable size on the server. Then, in KDT Recorder, format the drives again and confirm them for the software, and choose whether recording stops or overwrites old files when the disk is full. Next, in Device, add a device of type Recorder with the recording server's IP address, and add the cameras. Finally, in Record, choose each camera, the main or sub stream, and the hours and weekdays to record, and decide whether recordings go to the server's storage or to the device's own storage.

Events in KDT VMS run through an event server. In Device add a device of type Event SVR with the event server's IP address, assign the cameras to it under Event, then double-click each camera to enable the events you want, such as motion detection, which must also be switched on in the camera's own menu. In Preview tick Enable Alarm triggered Pop-out Image so that a separate window shows the picture when motion is detected, and use Alarm Center to watch live events and search past ones.

There are three ways to reach devices from outside, and they differ mostly in security. A VPN into your own network is the most secure and works for the VMS, the web page and the app alike; P2P (a cloud service) is the simplest for phones, because the device connects out to the cloud service and as a rule needs neither a public IP nor router configuration; port forwarding with DDNS is the most fragile and leaves the largest attack surface open. Use a VPN for administrators and VMS clients, use P2P for occasional phone viewing if you trust the manufacturer's cloud service, and treat port forwarding as a last resort. The specification tables of KDT NVRs and cameras list P2P and DDNS among their supported protocols.

An open port makes the device reachable by everyone: automated scanners find open camera and recorder ports quickly and try default and common passwords, the Mirai botnet of 2016 was built from cameras and recorders that were reachable directly from the internet and still used factory passwords, and outdated firmware adds known vulnerabilities on top. If you have no other choice, allow only HTTPS, use an external port other than the default one, strong unique passwords and current firmware, restrict the allowed source addresses on the router, and turn UPnP off so that devices cannot open ports by themselves.

Port forwarding only works if your router has a public IPv4 address. Many mobile connections and some fixed lines sit behind CGNAT, whose addresses are in the 100.64.0.0/10 range (RFC 6598) or in private ranges; if the router's WAN address is in such a range, or differs from what a 'what is my IP' website shows, no port forward can work, and DDNS does not fix it because it only maps a name to your current IP. Ask your internet provider for a public IP, or use P2P or a VPN through a server that has a public address.

KDT provides its Android and iOS apps in the download center. If the device shows as offline, check in this order that the NVR itself has internet access (correct gateway and DNS), that its P2P or cloud status is online, that its date and time are correct and that the password matches the one saved in the app. If it connects but is slow, open the sub-stream instead of the main stream, use Wi-Fi instead of mobile data and reduce the number of tiles on screen; a sub-stream typically needs about 0.5 to 2 Mbps, while a 4 MP main stream typically needs about 2 to 8 Mbps, depending on the codec and the scene.

Decoding load is measured in megapixels per second: channels × megapixels per picture × frames per second. For example, 32 cameras of 4 MP (2560×1440, about 3.69 MP) at 25 fps come to about 2,950 MP/s, while the same 32 tiles on 720×576 sub-streams come to about 330 MP/s, roughly 9 times less. For its KH-WSC714-32/84 video wall workstation (20-core processor, 32 GB RAM and 8 GB dedicated graphics), KDT publishes decoding ratings of 50 channels at 1080p, 25 at 4 MP or 12 at 4K, and all three work out to roughly 100 megapixels per frame, so count pixels rather than cameras. Use sub-streams for tiles and the main stream only for full-screen views.

Software decoding runs on the CPU and becomes heavy quickly as streams are added; hardware decoding uses the video engine of the processor's integrated graphics or of a graphics card and decodes many more streams with far less CPU load. KL-VMS+ lists hardware acceleration and load sharing with on-board or external graphics cards. At the same resolution H.265 needs more decoding power than H.264, so it depends more on hardware decoding: processors and graphics cards since roughly 2016 generally decode it in hardware, older ones do not. To check, open Task Manager, Performance, GPU and watch the Video Decode graph while the video wall is running: if it moves while the CPU stays low, hardware decoding is active.

Smearing and gray blocks usually mean network packets are being lost: where the VMS allows it, switch stream transport from UDP to TCP, avoid Wi-Fi for the client and the camera link, and check the switch port for errors. Stuttering and slowness are usually a question of volume, so add up the stream bitrates and compare them with the PC's network port: 32 main streams of 4 Mbps make 128 Mbps, more than a 100 Mbps port can carry but well within 1 Gbps. KL-VMS+ lists Adaptive Streaming to manage client load; lowering the frame rate of tiles, enabling hardware acceleration and updating the graphics driver also help.

The limit is the video outputs and the decoding power, not the software: KL-VMS+ lists no limit on the number of monitors (hardware only), and KDT's KH-WSC714-32/84 workstation drives four independent displays from one HDMI 2.1 and three DisplayPort 1.4a outputs, with up to 200 channels in total. Plan one output per monitor, and remember that a monitor showing four full-resolution 4K streams needs the decoding budget of 4 channels of 4K. Split very large walls across several workstations that each show their own group of cameras, and keep the operating system and the VMS database on an SSD.

Recording gaps on a Windows server usually come from the operating system rather than the camera: automatic restarts after Windows updates, sleep mode, antivirus scans locking the recording folders, or recording to the small system drive. Set Active Hours or a fixed maintenance window, choose the High performance power plan with sleep off, exclude the recording folders from real-time scanning, record to separate disks, keep the operating system and VMS database on a separate SSD as KDT does on its VMS Plus storage servers, and put the server on a UPS. Check the VMS system log for the exact time of each gap: if gaps always repeat at the same hour, look for a scheduled task or update.

Three things must be set. In BIOS or UEFI, set the power-loss recovery option (often called AC Power Recovery or Restore on AC Power Loss) to power on automatically; in the software, enable starting with Windows and opening in the last state, which in KDT VMS is under Config in Control Panel; and on a display PC, configure automatic Windows sign-in so that the login screen does not block the way. With a UPS, short outages do not interrupt recording, and during longer outages it allows a clean Windows shutdown.

Give every person a personal account and only the minimum rights needed: for example, operators get live view and PTZ, investigators get playback but not export or delete, and only administrators change settings. KL-VMS+ supports user groups, permissions per resource and camera, time-based access, limits on the recent archive interval a user can view, PTZ priority levels (so that a supervisor can override an operator, for example) and sign-in through Active Directory; in KDT VMS, users with different permission levels are created under Control Panel, Account. Avoid a shared administrator account because it makes the activity log useless, and remove each employee's access on the day they leave.

Yes. KL-VMS+ keeps a system log and an action history that can be reviewed by user and by action, and the KL-PRS and KL-FRS datasheets list real-time user and system logs that can be searched by date and user. Use them to review what each user did, for example who exported footage, who changed a camera's settings and when someone signed in. Export the log regularly and keep it at least as long as your video retention; it only means something if every person has a personal account.

Where the device supports several accounts or roles, create a separate account for the VMS on each camera and recorder instead of logging in with the administrator account, and give it only the rights it needs (live view, playback, PTZ, and settings only if you manage them from the VMS). If the VMS PC is ever compromised or a technician leaves, you change that one account rather than the administrator password of every device. Choose long, unique passwords (typically 12 characters or more), keep them in a password manager, and change every device's factory password before it goes on the network.

Start from the time window, not the camera: take the approximate time of the event with a few minutes of margin on both sides and open the cameras along the route together, because KL-VMS+ supports synchronized playback across cameras. Then narrow it down with event or motion search, smart search and the timeline thumbnails; in projects with smart analytics, searching by attributes such as color, age or plate is also possible. Mark what you find with a bookmark (date, title and description). Fast playback is the last resort: even at KL-VMS+'s maximum of 127x, one day (86,400 seconds divided by 127) takes about 680 seconds, about 11 minutes, and small events are missed at that speed.

Export in an open, widely playable format: MP4 with H.264 is the safest choice, an H.265 file needs a player with HEVC support, and a manufacturer's own format opens only in that manufacturer's player, so put the player on the media as well; KL-VMS+ lists export of single and grouped cameras in standard formats. Immediately after exporting, calculate the file's SHA-256 checksum (in Windows with certutil -hashfile filename SHA256 or the PowerShell command Get-FileHash), record it in the case report and keep the original recording on the NVR until the case is closed. Some software also embeds a watermark or signature, but a checksum recorded separately proves that the file has not changed since export. Format flash drives as exFAT or NTFS, because FAT32 cannot hold a file larger than 4 GB, and one hour of 4K video at 16 Mbps is 7.2 GB.

All cameras, recorders and the VMS server should take their time from the same NTP source (NTP uses UDP port 123 and is defined in RFC 5905), ideally an internal server that itself syncs with a reliable outside source. Evidence depends on it: if a camera's clock runs 3 minutes fast, an event that really happened at 14:00 is stamped 14:03 on it, so a search for 14:00 to 14:02 misses it and the order of events between cameras can look reversed. Keep the difference under one second and use the same time zone everywhere: Iran Standard Time, UTC+03:30.

Since 2023 (1402 in the Iranian calendar) Iran no longer changes its clocks and stays on UTC+03:30 all year, but devices with old firmware or an old time-zone list may still apply the former daylight saving rule (DST) and move the clock one hour forward in summer. On cameras, NVRs and servers, turn DST off or choose a fixed UTC+03:30 time zone, and check the time again after every firmware update. Footage recorded earlier keeps the wrong timestamps, so note the offset for the affected period so that it is interpreted correctly in reports and evidence.

Pixel-based motion detection reacts to any change: shadows, swaying trees, rain, insects around infrared lights and sudden changes in lighting. To reduce false alarms, draw detection zones that leave out trees, roads and reflective surfaces, lower the sensitivity until a person at the farthest point still triggers it, limit the schedule to the hours you actually need, and switch to smart events such as line crossing, area intrusion or human detection, which the specification tables of KDT NVRs list among their smart functions. Test after every change by walking through the scene in daylight and at night.

Choose by how critical the event is: a pop-up and sound in the VMS reach only someone who is watching the screen, email (SMTP) is delayed and can land in spam, mobile notifications depend on the phone's internet connection and the manufacturer's cloud service, while an alarm output (relay) to a siren or light works even when the internet is down. For important events use at least two channels, for example a VMS pop-up plus a mobile notification, and let the relay output drive a local siren. KL-VMS+ has an alarm rule engine that can run a defined scenario after an alarm and manages system messages in three categories: warning, notification and information.

A running program is not enough; monitor the state of every channel (connected, disconnected, recording). KL-VMS+ offers a health monitoring dashboard, system notifications, channel status and reports on each server's and camera's bandwidth as well as connection status, downtime and recording status over a chosen period, exportable to Excel. In addition, once a week play back a random camera at a random time to prove that the recording exists and plays, and connect camera-offline and disk-full conditions to alerts.

Update in a planned window, not during critical hours: read the release notes, first back up the configuration and database (and the license file), and usually update the server before the clients so that all of them end up on the same version. Test the new version on one client or a spare PC first, and keep the previous version's installer so that you can go back. According to their datasheets, KL-PRS and KL-FRS are updated with a compressed Zip file that needs no technical knowledge, and KL-VMS+ updates are free during the support period. After the update, check that all cameras are online, recording is running and the time is still correct.

Yes, through ONVIF: the specification tables of KDT NVRs and cameras list ONVIF (Profile S, G and T on all NVRs and most cameras, Profile S only on some cameras), KL-VMS+ supports ONVIF (as the licensed module KL-VMS/O+) and integration through vendor SDKs, and KL-PRS and KL-FRS accept cameras of any brand. Profile S covers live video and PTZ control, Profile G covers recording and playback from the device's own storage, and Profile T covers advanced streaming with H.265, imaging settings and events such as motion and tampering. Expect live view, PTZ and recording to work; brand-specific smart features often need a native integration, and the KL-PRS and KL-FRS datasheets likewise state more features with KDT's own cameras.

Automatic discovery of ONVIF cameras uses WS-Discovery multicast (UDP port 3702 to address 239.255.255.250), which routers do not pass between subnets or VLANs, so the VMS only sees cameras in its own network segment. If nothing is found, check that the PC and the cameras are in the same subnet, that the Windows firewall allows the VMS on this network and that the switch does not filter multicast traffic; otherwise add the cameras by IP address or by an IP range. To find devices and set their IP addresses you can also use the IP tools in the Windows software folder of the download center (KDT PRO Device Manager and IP Tool Finder); KDT VMS itself has a Search menu for finding devices and, for KDT devices, a device menu for changing the IP address.

The specification tables of KDT NVRs and cameras list compatibility with API, CGI, ONVIF (Profile S, G and T on most models) and SDK, and list RTSP among the protocols, so your software can use the documented interfaces or simply pull a stream over RTSP (the default RTSP port is 554). ONVIF is the best choice for compatibility across manufacturers because it is a vendor-neutral standard; use the SDK or CGI only when you need a function that ONVIF does not expose. According to its datasheet, KL-VMS+ can also be extended with customer-specific features and localized on request.

Analytics can run in the camera (Edge) or on a server or AI appliance (central). Edge analytics put no processing load on the server, keep working when the recorder is offline and send only events, but are limited to what the camera itself offers; central analytics are richer and can be updated in one place, but they must decode every analysed stream and scale with GPU or AI capacity. KL-VMS+ supports edge and central analytics by license, and typically not every frame needs to be analysed: analysing 5 to 10 frames per second at moderate resolution cuts the load a lot.

According to the specification tables, the KH-iES313-8/2X processes up to 2 license plate channels, the KH-iES513-16/6X up to 4 plate channels and 2 face recognition channels, and the KH-iES715-32 up to 8 face recognition channels and 16 perimeter protection channels with an additional license, plus 4 channels of advanced safety and behavior analytics with a Pro analytics license. The tables also name the installed software: plate recognition on the first model, plate and face recognition on the second and AI VMS PLUS on the third. These figures apply only to these devices; on your own PC the capacity depends on the GPU, the resolution and the analysed frame rate, so test with the project's real cameras before finalizing the design.

Everything is in the download section of the KDT website: ten folders for firmware, Windows software, license plate recognition (ALPR), face recognition, Android apps, iOS apps, user manuals, video samples, datasheets and catalogues, plus a search box for file names, product models or software names. Each folder opens in the KDT technical download center, and the KDT VMS manual (the file KDT-VMS.pdf) is in the Windows software folder under Vms. Download only from there, match the version to your device model and keep the installer of every version you deploy.

KDT's remote support uses KDT Desk, KDT's own remote-connection program, which is in the Windows software folder of the download center. Run the program and give the support specialist the ID and one-time password that it shows so that they can connect. This lets the specialist see the settings and status directly, and according to KDT's after-sales service description many problems are resolved without anyone visiting the site.