• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 98 subscribers
  • Views 550 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Splitting sensor image into 2 streams with DM385

Eli Balta
Intellectual 935 points


I am developing with IPNC RDK 3.5 and wish to split the captured sensor image into 2 separate regions, which will be encoded into two separate streams.

Is this supported by McFW? Has anyone succeeded in implementing something like that?

My guess is that I would somehow have to use a DUP link to duplicate the captured video, and then use a resizer link on each of the streams to crop the unneeded half.

Any information would be very welcome.

  • 0
    Guru 31670 points

    I'm not sure about the same in IPNC RDK using ISS. But in HDVPSS subsystem, you can get this done using multiple scalars. One using VIP scalar (SC_3) on the fly, and another using M2M scalar. But a simpler hardware solution is also possible, by sharing the same input with 2 VIPs simultaneously and using two VIP scalars (SC_3 and SC_4) in parallel so that you'll have two different streams, without much memory load of using M2M scalar. I think this a crude way still.

  • 0 in reply to Renjith Thomas
    Intellectual 935 points

    Thanks for your answer,

    Could you please provide a little more info regarding the hardware solution?

    What input should the VIPs be connected to? Does it matter which VIPs are used?

    Also, where could I find documentation regarding the various available scalars?

    Thanks,

    Eli

  • 0 in reply to Eli Balta
    Guru 31670 points

    Hi Eli,

    Are you still looking into the same issue? Sorry for my delayed response. 

    You can give the sensor input to VIP. VIP supports many formats except Bayer. It doesn't really matter which VIPs are used. 

    You can have a look at the DM814x TRM for high-level documentation. More information is only available from TI after signing an agreement. You'll have to contact TI sales representative for this.

  • 0 in reply to Renjith Thomas
    TI__Guru**** 486995 points

    Hi,

     

    I guess you want to encode different regions of the already captured image, so you could capture either from iss or vip, then extract portion of the image and encode it.

     

    Rgds,

    Brijesh

  • 0 in reply to Brijesh Jadav
    Intellectual 935 points

    Brijesh,
    How would this be implemented?

    Does any of the links available in McFW support image cropping, or would I need to implement a custom link that does this?
    The usecase as of now is pretty simple
    ISS Sensor -> Camera -> BitsOutVPSS -> BitsInVideo -> Encode -> BitsOutVideo -> BitsInHost -> RTSP stream
    The result I need is two separate RTSP streams, each containing a different region of the picture.