onvif-nvt

ONVIF library for NVT (Network Video Transmitter) devices.

This package is written with Javascript classes and ES6. Both Promises and Callbacks are supported. Note: Even though callbacks are supported, if you pass parameters incorrectly, you might end up getting a Promise back.

The ultimate goal of the onvif-nvt package is to have as much complete coverage of the ONVIF spec as possible.

The onvif-nvt package will work only with node server-side. Ultimately, you would use sockets to communicate between a client and server the desired ONVIF commands.

Installation

Note: This is an active project, and while so, there will be many updates. PRs are welcomed.

npm install onvif-nvt

Sample Application

You can find the sample application that uses Vue and Quasar here.

Contributing

Before making a PR please do the following:

• Make sure you are consistent with style.
• npm run lint-fix
• Make sure you have updated any functionality with JSDoc notations.
• npm run jsdoc
• Make sure you have added to run tests (live testing).
• If you have a specific camera not identified, create a folder for it in test/data/xml/{camera}. Create configuration data for it in run.config.js. Run all run tests with saveXml.setWritable(true) (in run.js) and capture all the xml.
• Make sure you have added to jest tests (uses the aforementioned xml). Update test/config.js with your camera.
• node run/run.js
• npm run test If all is well, make the PR.

Available Functionality

• Core (Device)
  • Discovery Mode, Device Information, System Date and Time, Scopes, Services, Capabilities, ServiceCapabilities, DNS, Network, Reboot, Backup, Restore, GeoLocation, Certificates, Relay, Remote User and many more.
• PTZ
  • Nodes, Configuration, Absolute move, Relative move, Continuous move, Geo move, Stop, Status, Presets, Home Position, and Auxillary Commands.
• Media
  • Profiles, Video/Audio/PTZ/Analytics/Metadata Configurations, Video sources, Audio sources, Stream Uri, Snapshot Uri, Multicast, and OSD.
• Snapshot
  • Using the snapshot module, you can retrieve snapshots from the device.
• Other modules
  • Access, Access Rules, Action, Analytics, Credential, DeviceIO, Display, Door, Events, Imaging, Media2, Receiver, Recording, Replay, Schedule, Search, Security, Thermal and Video Analytics.

The library is made in such a way that only modules that will work with your device are automatically included. For others, you can choose whether or not to bring in the functionality.

Example (Discovery)

const OnvifManager = require('onvif-nvt')
OnvifManager.add('discovery')
OnvifManager.discovery.startProbe().then(deviceList => {
 console.log(deviceList)
 // 'deviceList' contains all ONVIF devices that have responded.
 // If it is empty, then no ONVIF devices
 // responded back to the broadcast.
})

Example (Connect and Continuous Move)

const OnvifManager = require('onvif-nvt')
OnvifManager.connect('10.10.1.60', 80, 'username', 'password')
  .then(results => {
    let camera = results
    if (camera.ptz) { // PTZ is supported on this device
      let velocity = { x: 1, y: 0 }
      camera.ptz.continuousMove(null, velocity)
        .then(() => {
          setTimeout(() => {
            camera.ptz.stop()
          }, 5000) // stop the camera after 5 seconds
        })
    }
  })

Example (Snapshot)

const OnvifManager = require('onvif-nvt')
OnvifManager.connect('10.10.1.60', 80, 'username', 'password')
  .then(results => {
    let camera = results
    // calling add method will automatically initialize snapshot
    // with the defaultProfile's snapshotUri
    camera.add('snapshot')
    camera.snapshot.getSnapshot()
      .then(results => {
        let mimeType = results.mimeType
        let rawImage = results.image
        let prefix = 'data:' + mimeType + ';base64,'
        let base64Image = Buffer.from(rawImage, 'binary').toString('base64')
        let image = prefix + base64Image
        // 'image' is now ready to be displayed on a web page
        // ...
      })
    }
  })

Example (Events)

PullMessages now works with Hikvision, TrendNET and Pelco. They are not working with Axis.

const OnvifManager = require('onvif-nvt')
OnvifManager.connect('10.10.1.60', 80, 'username', 'password')
  .then(results => {
    let camera = results
    // if the camera supports events, the module will already be loaded.
    if (camera.events) {
      camera.events.on('messages', messages => {
        console.log('Messages Received:', messages)
      })

      camera.events.on('messages:error', error => {
        console.error('Messages Error:', error)
      })

      // start a pull event loop using defaults
      camera.events.startPull()

      // call stopPull() to end the event loop
      // camera.events.stopPull()
    }
  })

Documentation

onvif-nvt documentation

Testing

Note: The code for testing is not available in the npm package.

All functionality has been tested with Hikvision (fixed and ptz), Pelco (ptz), TrendNET (fixed) and Axis (ptz).

Functional Testing

Functional testing is intended for 'live' testing with an actual ONVIF device. It is done with the run.js in the run folder. This will test the actual capabilities of a camera, but just as importantly, it can capture the xml response to file so it can be used in Jest tests. The configuration file for the run tests can be found in run/config.js.

For discovery testing, just set the runDiscovery variable at the top of the file to true.

For core testing, set the runCore variable at the top of the file.

You can do this for other modules as well. See file for options.

Set up your local camera by setting the appropriate variables in run/config.js, like this:

const address = '10.10.1.60'
const port = 80
const username = 'root'
const password = 'root'
const folder = 'hikvision'

Run the run.js file via node (personally, I use VS Code - an amazing editor/debugger).

The tests will only run onvif modules supported by the camera capabilities, so if it doesn't support PTZ, then the PTZ tests won't be run. Same for other modules.

Your mileage may vary as I have found some ONVIF devices don't support very basic functionality, like GetSystemDateAndTime. In some cases, you might get a response back of Action not supported. If this happens, then the tests will fail. This is nothing to be alarmed about. You may also get other errors. The most common I see is if your camera does not support audio configurations. Also, you may get a Not Implemented error, meaning the implementation of that method has not yet been added.

Automated Testing

Jest is being used to do the automated testing and code coverage. All tests are in the tests folder, as well as XML Requests and Responses from various ONVIF devices. To start the tests, run the following: npm run test

Inside the test folder is a config.js file where you can set options for the tests.

Request

onvif-nvt uses request for device communication. Some people may believe this is too heavy of a package and that http should be used instead. request works quite well with HTTP digest and digest realm scenarios. In a lot of cases, this is needed for snapshot access. However, I am finding some ONVIF devices that allow you to turn on HTTP security (besides the ONVIF user token security) which provides a double layer (albeit, with the same username and password). It is these cases where request just works.