While running object detection example for Hailo (examples/hailo/detect.py) with the image of an orange, it detects it somewhat poorly, detecting only a vase sometimes:
When I changed the format in lores to BGR888 instead of RGB888, it detects an orange very nicely (which it wouldn't if the BGR format was incorrect, because the orange would be blue):
I can only conclude that hailo.run() or models themselves expect the image in the BGR888 format instead of RGB888. Is that correct?
Hardware :
Raspberry Pi 5 8GB with Arducam IMX708 B0310
Yes, I think that's usually the case. Of course it could vary with the model, but I think pixels in R, G, B order in memory (which libcamera slightly strangely refers to as BGR888) is most common.
While running object detection example for Hailo (
examples/hailo/detect.py
) with the image of an orange, it detects it somewhat poorly, detecting only a vase sometimes:When I changed the
format
inlores
toBGR888
instead ofRGB888
, it detects an orange very nicely (which it wouldn't if the BGR format was incorrect, because the orange would be blue):I can only conclude that
hailo.run()
or models themselves expect the image in theBGR888
format instead ofRGB888
. Is that correct?Hardware : Raspberry Pi 5 8GB with Arducam IMX708 B0310