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.

TDA4VM: How do I add a new tiovx node to separate interleave frames to different buffs

Part Number: TDA4VM

TDA4VM EVM board

SDK 0804

Please refer to the post below for details

TDA4VM: if ub960 work in Line-Interleave mode and with 4 camera , how to separate 4 camera image into separate buffers - Processors forum - Processors - TI E2E support forums

Now we want to create a new node to split the image, how do we do that

  • Hi,

    Please refer to below documentation for adding a new node in tiovx.

    TIOVX User Guide: User Kernels and User Target Kernels

    Regards,

    Broijesh

  • Hi Brijesh,

    The customer Hirain is using DMA to split the frame, but their current image data is 40ms per frame, 12MB per frame, including four pictures. They are currently testing the operation of splitting images through dma on MCU2_0. The test time for pure dma operation is approximately 6ms for each picture, which takes up a total of It’s 24ms, and the rate is 500MB/S. If it looks like this, the occupation of DDR bandwidth and the blocking time for subsequent image processing are not very reasonable. Do you have any optimization suggestions? 

    Xu have proposaled the 4D DMA, it can do this split in one time, but we need your help to implement it.

    BR,

    Biao 

  • Hi Biao,

    As we discussed, can we please try to understand why customer is not using different vc id in ub960? Because if this is done, then CSIRX can separate them in different buffers and we dont need to run any additional path to separate them

    Regards,

    Brijesh