In the docs sensor.snapshot() states ’ Takes a picture using the camera and returns an image object.’
I wonder if there a comparable function to read_into() bytearrays/memory views.
Wouldn’t that be beneficial and safe allocations and copying and in general what one would prefer?
Also, to keep up the camera frame rate we have a triple buffer architecture where a frame is always being captured continuously in the background. When you call snapshot() you are just kicking off the DMA system (if not running), and then grabbing whatever the latest frame is (or waiting for it).
Note that all images are buffer objects so getting the bytes() and using memory views on them work.