Projets-2016-2017-UPnP Cameras integration into OpenHAB

From air
Jump to navigation Jump to search

Team

  • Supervisor: Didier Donsez
  • Members: Antoine Blanc, Christopher Law

Calendar

  • Start of the project : January 9th, 2017.

Git repository

 repository with UPnP Camera binding

Week 1 (January 9th - January 15th)

Week 2 (January 16th - January 22nd)

Week 3 (January 23th - January 29th)

  • Installation of Kodi and Apache Felix
  • First try of UPnP protocol
  • Connect cameras to local network: D-Link cameras can only be configured with a computer running on Windows or Mac and not Linux because of the software provided by D-Link. However, there is still a solution for Linux user by using wine.
  • Cameras used in the project have to be configured with the following software:
 DCS-5222L: (A version): Windows user and Mac user
 DCS-932L: (A version): Windows user and Mac user

 The network must allow multicast.
 An Ethernet cable is needed for the first steps of the configuration.
 Reset cameras when using them for the first time in order to set a password.
  • Detection of other devices through UPnP with Felix and Kodi: repository

Week 4 (January 30th - February 5th)

  • Using openHAB2 through the Karaf console can be time consuming with the exportation of bindings and find the different commands to install and activate them. Therefore, we decided to install and run openHAB from Eclipse IDE.
  • Set up of Eclipse IDE for Eclipse SmartHome developments and openHAB2 addons

Week 5 (February 6th - February 12th)

  • Installation of UPnP binding in OpenHAB
  • Detection of camera through OpenHAB interface with UPnP protocol
  • modification of sitemaps in openHAB

Week 6 (February 13th - February 19th)

Week 7 (February 20th - February 26th)


Week 8 (February 27th - March 5nd)

  • Continue the clean creation of the binding
  • Coding guidelines
  • Fix errors and typos on the binding
  • Find documentations on motion tracking

Week 9 (March 6th - March 12th)

  • Creation of a new branch upnpcamera.
  • Started implementation of motion detection with openCV and Java. Some problems with the installation of openCV.


Week 10 (March 13th - March 19th)

  • Improved motion detection algorithm.
  • Implementation of the new binding.


Week 11 (March 20th - March 26th)

  • Matched motion detection with camera tracking.
  • Modification on the new binding.
  • Made tests to control camera and motion detection.

Week 12 (March 27th - April 2nd)

  • New branch created to fit the new binding name DlinkUpnpCamera
  • Started pull request: Pull request link
  • Wrote report and presentation slides