• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 98 subscribers
  • Views 258 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.

How does DM355 process video signals coming from multiple sensor simultaniously.

Myung-Jin Sung
Prodigy 10 points

Hi everyone,

I'm searching how DM355 gets two camera input simultaniously.
but, I can't get a sense how it works although I read the reference-guide for DM355.
we know this function is supported, as we see the description like below in reference-guide.

In the reference-guide...
'Support for progressive and interlaced sensors (hardware support for up to 2 fields and firmware
support for higher number of fields, typically 3-, 4-, and 5-field sensors)'

Below is my questions, please give me the informaiton.
1. how DM355 gets two camera input simultaniously.
2. then how it records the two input.
3. if we display two camera input on one LCD, how it works.
4. please let us know the work flow and any references like source code.
5. for using two camera input simultaniously, does we have any cosideration like below.
 - is it possible to use the different interface (e.g. one is raw, the other is BT656)
 - or does we use the same interface (e.g. one is raw, the other must be raw interface)
6. is any method  to test the these functions.
 - It's possible to test these funtions by using DM355EVM board with composite and raw camera sensor interface.
   eq. one is composite sensor(generic Ycbcr interafce), the other is cmos sensor (raw interface)

please reply soon. thanks.

  • 0
    TI__Mastermind 37340 points

    I will do my best to answer your questions.  First, I should point out that as it pertains to progressive vs. interlaced, in the interlaced case the original video frame is broken up into two or more fields (2 being the most common); this does not mean you have more than one input, but rather that your one input is broken down into several fields allowing for faster refresh rate at a lower data-rate (useful when data cannot be trsaferred fast enough or lower data-bandwith is desired)

    now to answer your questions

    1. I believe the only way to do this is to use external parts that can consolidate different video inputs into a single video input (Bt.656, YCbCr or RAW) for DM355.  Here I can see two choices, you can have a time-miltiplexed input in which each video input is represented by one frame or field multiplexed in time or you can reduce the size of all video inputs and have all of them appear in a single video frame (sort of the security camera model).

    2. If you are inputing the 5 frames in a time multiplexed manner where each field belongs to a separate video input (with the help of externel part), then you can write one frame at a time into a file or use a separate buffer (or two) for each video input.  Your application will need to keep track of which buffers belog to which inputs and either store them to the corresponding file or display them.  If you are considering the resizing down approach, then you can always resize back up the portion of the frame that makes up the video input of interest.

    3. Most likely, you will need to resize both inputs into single output frame (security camera model where multiple inputs are shown in one screen).

    4. I am not aware of any reference source code for this.  As I suggested, if you use an external part to accomplish security camera model of multiple inputs, from a DM355 perspective it will be treated as single input so the existing drivers should work.  Any manipulation of the video data will be done by your user space application.

    5. You can only select one type of input at a time, so either BT.656 or RAW but not both.

    6.  As I said in 5, only one input type at a time; if you adhere to this, you should be able to test this with our DM355 EVM; composite input is built into EVM for, camara RAW input, you will need to use DC5 connector and probrably attach a daughter card with your desired camara sensor.