Hi all:
we capture two 1080P 25fps video, and on dsp we analysis the video frame characteristic,
and we the perspective transformation matrix, but when we do the perspective transformation to stitch the two frames,
it seems too many pixels to render, it costs too much time,
my question is can we use dm8168 SGX530+OpenGL to do the render, dsp just do analysis characteristic and get perspective transformation matrix,
or some else good ideas?
best regards
xavier
