Basics of Simultaneous data and video acquisition: please help!

[groomit]

Hi everybody,

I am using Labview 2017 and a NI USB-6343, X Series DAQ Device on a Win10 PC for continuous acquisition of 12 potentials at 1024 Hz for 24-48 hours. I would now need to add continuous video recordings "synchronized" (say, within 1/50th of a second or so) to the acquired data. Ideally, I should end up going through my recordings off line, locating points in time with given patterns of potentials, and then checking what happened on the video in the few seconds before and after that point in time. I am used to performing off-line analyses with Matlab, have some experience with data acquisition with Labview, whereas I am naive concerning video acquisition and processing. I have already purchased a Basler ACE 1300 60 gm NIR camera and a separate PoE adapter. My PC is equipped with a Realtek gaming GBE controller.

I would greatly appreciate suggestions and hints regarding the following questions:

1) Is it feasible to modify the VIs I have already programmed adding the acquisition of video recordings with a time stamp, compression of the video recordings to save disk space, and disk storage of one video file every hour in addition to the usual data files?

2) I understand I should buy the Vision Acquisition Software, whereas I would not need the Vision Development Module. Is that right?

3) Should I consider buying also a GigE frame grabber?

Thanks in advance for your advice and best regards

[Bob_Schor]

If you are going to use LabVIEW for something like this, I would definitely advise getting both the VDM and VAS.  Were you planning on obtaining continuous Video, or episodic Video based on a "feature" of the recorded signals?

We had a project recording behavior in mice.  We had one system that recorded a channel of data at 10 Hz, along with a Video (640x480, color, 30fps) signal.  When an interesting "event" occurred on the data channel, we saved an AVI made from the Video signal from 1 second before the event to 3-4 seconds after.  We also recorded an "Event" channel that noted the time (measured in milliseconds from the beginning of the Recording Session) that the Video started and stopped, with the Data Channel(s) sharing a similar Time encoding.  This simplified the post hoc analysis of the Data records, including locating the "event" signals in the Analog record and matching it with the corresponding AVI.

This was all done in LabVIEW, facilitated by LabVIEW's Data Flow Programming model that allows multiple processes to run essentially in parallel.

Bob Schor

[groomit]

---

@Bob_Schor  ha scritto:

If you are going to use LabVIEW for something like this, I would definitely advise getting both the VDM and VAS.  Were you planning on obtaining continuous Video, or episodic Video based on a "feature" of the recorded signals?

We had a project recording behavior in mice.  We had one system that recorded a channel of data at 10 Hz, along with a Video (640x480, color, 30fps) signal.  When an interesting "event" occurred on the data channel, we saved an AVI made from the Video signal from 1 second before the event to 3-4 seconds after.  We also recorded an "Event" channel that noted the time (measured in milliseconds from the beginning of the Recording Session) that the Video started and stopped, with the Data Channel(s) sharing a similar Time encoding.  This simplified the post hoc analysis of the Data records, including locating the "event" signals in the Analog record and matching it with the corresponding AVI.

This was all done in LabVIEW, facilitated by LabVIEW's Data Flow Programming model that allows multiple processes to run essentially in parallel.

Bob Schor

---

Hi Bob,

thank you for your suggestions. I plan to obtain and save continuous video. Is VDM needed also for this? I ask because VDM is much more expensive than VAS, so I need to justify my budget. How did you manage video compression? Did you use a frame grabber?  

Best regards

[Bob_Schor]

We used an Axis M10 series camera (I don't remember the exact model, but I have a pair of them at home).  They are TCP/IP cameras that can be used with IMAQdx (and no additional hardware -- just camera connected to PC via TCP/IP, either wired or even wireless).  We used the default AVI codec for compression.

I have to confess I'm a little unclear what parts of the Vision and Motion Palette belong to VDM and which to VAS.  I'm pretty sure that the IMAQdx stuff is VAS (as I once did an installation and forgot to Activate the VAS license, so I saw lots of Image Processing VIs, but no Acquisition). 

OK, Good News.  I went in and deactivated my VDM License to see what "broke" in my Project.  I had a utility that make a Checkerboard Image (for some tests) -- that broke, but all the Image Acquisition and AVI stuff was OK.  Fortunately, I got everything back when I reactivated the license.

Bob Schor

[groomit]

Thank you!

[Jacobson-ni]

Also remember that your camera allows for hardware triggering which might be useful for capturing specific events.

[groomit]

… I have just finished installing the VAS -- it seems all AVI VIs actually require the VDM! So disappointing… Any suggestion on how to proceed? Perhaps saving individual frames and then doing postprocessing and AVI file generation in Matlab? 

[Bob_Schor]

If you are going to go the route of saving individual frames and then using "some other software" to make the Video, aren't there a number of routines, either free or inexpensive, that can do this sort of processing for you?  I would think that would be much faster and more efficient than Matlab, and maybe even than LabVIEW.  A significant virtue of generating the AVI in LabVIEW, of course, is you are directly saving a compressed file ...

Bob Schor