raufer92's profile - activity

Hi, i'm working in an image processing task with OpenCV (pyhton). I'm processing a video to detect/track some LEDs information (position, color, etc,...)

Here is a sample of a original frame,

After filtering, currently the frames look like this:

What i want is to detect all the leds, like illustrated in the next figure

Can anyone help me with some robust ideas to a general case? Thanks