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.

DM814x ISS Horizontal Image Offset ISS 3.5

Hi folks,

The configuration that I am using has an image sensor connected to the ISS camera input on he DM814x.  The sensor is also connected to another device on the board that captures the sensor image too (so effectively there are two paths from the sensor).

What I am seeing is that the video frames captured on the ISS are horizontally offset to the left by approximately 100 pixels.  This results in the captured frames losing a column of about 100 pixels wide on the left of the image and gaining a black vertical strip about 100 pixels wide on the right of the image.

It is almost like the horizontal sync signal is arriving too late for the video image.  However, capturing an image from the other device on our board that receives the image from the same sensor at the same time shows that the horizontal and vertical sync signals seem to be OK.

I am using ISS software version 3.5 combined with EZSDK 5.05.

Has anyone come across this behaviour before?

Is there any way in software to adjust when an ISS image capture starts?

Thanks in advance,

Regards,

Terry

  • Terry,

    Can you try to invert the HSYNC polarity and see what happens? 

  • Hi Renjith,

    Thanks for the quick reply.

    I've tried inverting the sync polarity at both the ISIF and IPIPEIF levels (although not both at the same time). This was done by writing to registers directly rather than changing any initialization code.

    Modifying the polarity at the ISIF level resulted in:

    1 . The 100 pixel black vertical line moved from the right hand side of the screen to the left hand side.

    2. The 100 pixel image block that was missing on the left of the image is still missing (i.e.the image looks exactly the same except it is shifted right rather than left) .

    Modifying the polarity at the IPIPEIF level resulted in:

    1. What looks like about a 400 pixel black vertical line appeared at the left hand side of the image.  This would be consistent with the horizontal blanking area in the sensor image.

    2. The 100 pixel area that was missing on the left of the image is now visible but, it looks like about 400 pixels have gone missing on the right of the image.

    In both cases the ISS freezes which I suspect is caused by writing directly to the registers.

    Is there a single place in the ISS source code where it is possible to change the polarity of the sync at all levels within the ISS hardware?

    Thanks again for your help,

    Regards,

    Terry

  • The ISS driver polarities for different modules are set in different functions. However you'll be able to pass the parameters from a single place, most likely. I believe your problem is related to some ISS parameters only.

  • HI Neill,

          I feel some issue still exist with ISS settings/sensor settings. Please go through the sensor register  and make sure resolution & blanking what you have set is correct. Also from the ISS side look at the ISIF register settings like start position. Hope your ISIF->SPH register is set properly.

  • Hi folks,

    Just an update on this problem.  The problem was eventually traced to a set of values in the function  issIpipePreviewParamsInit() in the iss file iss_drv_ipipe.c.

    Fixing these values for the sensor I am using removed the image offset.

    Thanks for everyone's help,

    Regards,

    Terry