Difference between revisions of "WICE video"
Robin.karhu (talk | contribs) |
|||
| Line 14: | Line 14: | ||
After the cameras have been installed and configured, a video task needs to be created in the WICE Portal. | After the cameras have been installed and configured, a video task needs to be created in the WICE Portal. | ||
When creating the video task, the trigger signal determining when recording of video should be started and stopped is configured. Available options are ‘Trigger button’, ‘Signal Reader’, 'Expression', and ‘None’. The trigger button option requires a physical button to be connected to the WCU with cameras connected, and will start/stop video recording when the user pushes the button. If trigging from Signal Reader is selected, the name of a signal recorder in a running Signal Reader | When creating the video task, the trigger signal determining when recording of video should be started and stopped is configured. Available options are ‘Trigger button’, ‘Signal Reader’, 'Expression', and ‘None’. The trigger button option requires a physical button to be connected to the WCU with cameras connected, and will start/stop video recording when the user pushes the button. If trigging from Signal Reader is selected, the name of a signal recorder in a running Signal Reader task must be supplied. When the trigger condition of the signal recorder is met, and the signal recorder started, the video recording will also be started. Alternatively, if "auto" is entered instead of a recorder name, the recorder name containing the WICE-internal signal Video_Camera_1 or Video_Camera_2 will be configured automatically. When the stop condition is met, the recording will be stopped. Similar to the Signal Reader trigger option, the 'Expression' trigger lets the user specify start and stop trigger conditions based on logical expressions containing signals measured by the Signal Reader module. Note that the user must make sure the signals present in the expression are actually measured (otherwise the expression will never be true). Unlike the 'Signal Reader' trigger, there does not have to be a recorder configured in the Signal Reader measurement assignment. | ||
The video module also supports recording of a pre-trig buffer, which means that video will be recorded for a configurable time period before the recording start event. If no triggering mechanism is selected (i.e. ‘None’), recording of video will be started as soon as the WCU has finished booting, and will be stopped when the WCU shuts down (typically at ignition-off). | The video module also supports recording of a pre-trig buffer, which means that video will be recorded for a configurable time period before the recording start event. If no triggering mechanism is selected (i.e. ‘None’), recording of video will be started as soon as the WCU has finished booting, and will be stopped when the WCU shuts down (typically at ignition-off). | ||
Latest revision as of 15:24, 29 November 2021
Introduction
The WICE video module makes it possible to record and monitor live video from cameras connected to in-vehicle WCUs. Video can be very useful in many automotive R&D scenarios for recording conditions in and around the vehicle, including user behavior, traffic situations, etc. The video module thus extends traditional sensor data capture modules with the ability to capture video synchronized with other time series data. Similar to other data capture modules in WICE, for instance the Signal Reader module, the video module allows both recording and monitoring of video signals, with recording triggered by configurable events. Monitoring in this context means live streaming of video from the WCU to a remote video playback tool. Recording of video is done to the SD card in the WCU, with recorded clips subsequently uploaded to the WICE back-end.
Configuration and set-up
To start using video in WICE, you need to connect one or two cameras to a WCU. The cameras are connected to USB ports on the WCU. Suggested and well-tested cameras are:
- Logitech HD Webcam C270
- Logitech HD Webcam C310
- Logitech Pro Webcam
- Logitech Webcam C930e
- Microsoft LifeCam HD-3000
The video cameras are configured in the WICE Portal module configuration for the WCU, reachable from the “Vehicles” tab.
After the cameras have been installed and configured, a video task needs to be created in the WICE Portal.
When creating the video task, the trigger signal determining when recording of video should be started and stopped is configured. Available options are ‘Trigger button’, ‘Signal Reader’, 'Expression', and ‘None’. The trigger button option requires a physical button to be connected to the WCU with cameras connected, and will start/stop video recording when the user pushes the button. If trigging from Signal Reader is selected, the name of a signal recorder in a running Signal Reader task must be supplied. When the trigger condition of the signal recorder is met, and the signal recorder started, the video recording will also be started. Alternatively, if "auto" is entered instead of a recorder name, the recorder name containing the WICE-internal signal Video_Camera_1 or Video_Camera_2 will be configured automatically. When the stop condition is met, the recording will be stopped. Similar to the Signal Reader trigger option, the 'Expression' trigger lets the user specify start and stop trigger conditions based on logical expressions containing signals measured by the Signal Reader module. Note that the user must make sure the signals present in the expression are actually measured (otherwise the expression will never be true). Unlike the 'Signal Reader' trigger, there does not have to be a recorder configured in the Signal Reader measurement assignment. The video module also supports recording of a pre-trig buffer, which means that video will be recorded for a configurable time period before the recording start event. If no triggering mechanism is selected (i.e. ‘None’), recording of video will be started as soon as the WCU has finished booting, and will be stopped when the WCU shuts down (typically at ignition-off).
Accessing recorded video clips
Video clips recorded on a WCU in a vehicle are uploaded at ignition-off to the WICE back-end and accessible through the WICE Portal web-GUI and through the WICE M2M REST API. Since video clips can be quite large, it can sometimes require more time to upload them than what is available at ignition-off. In this case the upload of the video clip will be resumed next time the WCU is booted. Uploaded video clips can be accessed through the WICE portal web-interface by selecting the Video task in the ‘Tasks’-tab and clicking ‘Get result files’, which will display a list of recorded video clips, that can be downloaded and played back.
Video monitoring
Monitored video streams are transmitted from the in-vehicle WCU over the mobile data communication link to an RTP video reflector. By connecting an RTP-based video tool to the reflector (e.g. Alkit Confero), one or more users can monitor the live video stream originating from the WCU.
Since many video-equipped WCUs can be transmitting video at the same time to the reflector, the video monitoring clients must select which video streams to receive. This is done using session set-up signaling between the client and the reflector, whereby the desired video source is identified using the WCU identifier as the session name.
Video codecs and file formats
Currently, two video codecs are supported in the WICE system: Motion-JPEG and VP8. These two codec represents two different trade-offs between bandwidth and computational complexity, where JPEG is low complexity / high bandwidth and VP8 is high complexity / low bandwidth. VP8 gives a much higher overall quality to bandwidth performance, but at the cost of considerably higher processing requirements, so depending on the WCU platform’s CPU performance, it might nevertheless be preferable to use JPEG to get a higher frame rate. Moreover, the recommended USB cameras support hardware JPEG compression, which further improves performance in terms of frame rate, especially at high spatial resolutions.
Two video file formats are supported: QuickTime and Matroska, with QuickTime being used for JPEG-compressed video and Matroska for VP8-compressed video (also referred to as ‘webm’ format). The webm-format has the advantage of being supported in HTML5, which means that video clips uploaded to the WICE Portal can be directly played back using a Portal-integrated video playback mechanism, whereas the QuickTime format requires a web-browser plug-in for streaming playback, or, alternatively, needs to be downloaded before being played back in e.g. Windows Media Player or VLC.
When video clips are uploaded to the WICE portal, clips from the same shot are concatenated and converted to MPEG-4, which makes it possible to play directly in the portal GUI.