Discrepancy viewing vs processing raw file

Discrepancy viewing vs processing raw file

The code below is meant to produce 2 black and white images, one for positive events and one for negative events

event_iterator.reader.reset() # Reset the reader
event_iterator.start_ts = int(1e6)
event_iterator.delta_t = int(1e6)

imgp = np.zeros((height, width), dtype=np.uint8) # Create an empty image
imgn = np.zeros((height, width), dtype=np.uint8) # Create an empty image
for evts in event_iterator:
imgp[evts['y'][evts['p'] == 1], evts['x'][evts['p'] == 1]] = 255
imgn[evts['y'][evts['p'] == 0], evts['x'][evts['p'] == 0]] = 255
break # Only do one iteration
return imgp, imgn

When I view the raw file I'm iterating above in metavision studio, I can see a decent number of positive and negative events in the range between 1s - 2s when the accumulation time is set to 100ms (see screenshots). I would then expect that I would be able to see two fairly populated images when processing them as above since my accumulation time is an order of magnitude larger than the accumulation time set in the metavision studio and I can plenty of events there.

However, this isn't the case. When I investigate the polarities in my events I see a large discrepency:

len(evts[evts['p'] == 0]) # number of negative events: 33
len(evts[evts['p'] == 1]) # number of positive events: 616

What am I misunderstanding about the way the events are displayed in the metavision studio vs the way they are processed using the event iterator? I don't want to use any frame generation algorithms, I need to be able to process the raw events using my own code.



    As a Prophesee customer, join the community conversation. 
    Request your free access today.

      Information

      In this Community Forum, you can discuss products, technology and applications. Only registred users can post but everyone can read. To ask a private question, enter a support ticket in My Area https://support.prophesee.ai/portal/en/newticket