• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 98 subscribers
  • Views 407 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.

Window masking in DM365 IPNC

Sudheesh
Expert 1945 points

Hi,

I am trying to add window masking (a rectangular portion of the video is always masked) feature to IPNC ( v1.5 ) in addition to face masking feature. My logic is this - Apply window masking before face detection logic runs. 

To this end, I added a 'DRV_frameCopy' function call to the function 'VIDEO_fdTskRun' in the file 'videoFaceDetectThr.c ' . I am using the virtual address provided to the variable 'pBufInfo' in the function 'VIDEO_fdTskRun' by 'OSA_bufGetFull' function . But when I use the virtual address as argument to 'DRV_frameCopy' , video is garbled. 

 

1. My function call - VIDEO_windowMasking(pBufInfo);

2. Assigning the Virtual address to be used in -DRV_frameCopy' -> frameCopy.dstVirtAddr = pBufInfo->virtAddr + VIDEO_BUF_HEADER_SIZE;

3. Calling 'DRV_frameCopy' function - DRV_frameCopy(NULL,&frameCopy);

 

Can you please tell me what I am doing wrong?

Sudheesh

  • 0
    TI__Mastermind 19985 points

    Hi,

    Logically, your code seems to be correct. I would recommend that you refer to alg_swosd.c file in the source tree (framework\alg\src). If you look for "(privacyMask==TRUE)" condition, you will be able to see the code which is used to draw the privacy masked rectangles. You can call the same API "ALG_swosdDrawRect()" from your application code with the correct set of parameters and it will draw the masked rectangle.

    Regards,

    Anshuman

    PS: Please mark this post as verified, if you think it has answered your question. Thanks.

  • 0 in reply to Anshuman Saxena
    Expert 1945 points

    Anshuman,

    Thank you very much for your reply. But as I have explained, I am using the same function (kind of) 'DRV_frameCopy' which is used inside 'ALG_swosdDrawRect' function. I created an instance of 'DRV_FrameCopyPrm' and filled in all the parameters. But when I use the virtual address supplied by ''OSA_bufGetFull'' function, the video is getting garbled. Is there any difference with the address returned by 'OSA_bufGetFull' and what is inside 'ALG_SwosdRunPrm *bufInfo' ? 

    Regards

    Sudheesh