Image processing with Theta V


#1

Hi,

I’m quite new about images generated by 360-degree camera. Can anyone give me a hint about how to do image processing on 360-degree image? (such as facial detection).

  1. For me, before thinking about image processing, it should be flat the 360-degree image into a flat panoramic image.
    From the post, it requires an external application Affinity Photo, and we need to manually crop the image.
    Is there any way can flat the image in code? Because I want take picture -> flat image -> image processing finished automatically (not manually)

  2. Or, without flattening the image, can I just assign some target point and process that specific area?


#2

People use OpenCV. If you type “opencv face detection” into Google, you will get a list. If you doing this inside the camera, Android has Face detection.

The output of the THETA S and V is an equirectangular image.

If you want to output dual-fisheye, you need to use a plug-in.

The facial detection will work on both dual-fisheye and equirectangular.

If you want to manipulate the equirectangular in your program, this may help:

http://theta360.guide/thetamedia/#_sample_application