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.

Omap4 blaze ducati camera ov5650 grain noise

Hi there,

I'm using 4AJ-2.5SP JB release on blaze board and have access to Ducati source code.

When testing with ov5650 back camera on blaze board [ also on custom board which has ov5650] I can see the grain noise in the preview as well in the recorded video with default settings.

Attached is the snapshot and the recorded video for reference.

What can be done to enhance the quality of camera output.

Regards

Anuroop

 [

2804.VID_20141009_180948.mp4

  • Hi Anuroop,

    This is most likely a problem with sensor settings inside the OV5650 driver. Check its setting & compare them against the recommendations in sensor's datasheet.

    Best Regards,

    Yordan

  • Hi Yordan,

    When I checked in the ducati driver for OV5650 we have gain table which is used ov5650_setExpGain() routine when sensor gain index is 300 (value [0] =112 [1]=256) this is when the grain noise is high.

    I have observed that the gain set is low if the lighting conditions are good  and when we have dark spots or object in ROI the exposure gain is increased and I feel the noise is also amplified along with signals.

    Can you explain what is the significance of the gain table and why it is like this?[attached spreadsheet

    2677.Again_table_ov5650.xlsx]

    I checked the sensor settings I did not see any specific issue with settings, however I can check again If I missed something.


    Regards

    Anuroop

  • Like any camera, OMAP4 camera is prone to noise in lowlight conditions. You can tune auto exposure or enable a noise filter depending on your preferences. For details please contact your TI representative.

  • Hi Mayank,

    Thanks for your Input, sure I'll check with our TI representative.

    In addition to earlier observations,


    1. Under Normal light conditions: Noise is present but less visible. [exposure time and gain is low]

    2. Under Normal light but focused on dark objects: Noise is more visible [exposure time and gain is high]

    3. Under Low Light conditions: Noise is prominent [exposure time and gain is high]

    I feel the noise is present all the time but amplified by the increase in exposure time and gain in ov5650 driver, I assume exposure time and gain is computed based on the luminance computation of past captures.

    Meanwhile till I get response from TI, if noise filters are part of ducati code base I can dig in to see if I can do something from my end.

    Regards

    Anuroop