At first, the interface contained only one function: to
process video frames by receiving a copy of the image data
and related parameters. As passing a multichannel high
resolution image through the interface as actual values, and
thus copying it into another memory location, caused a
significant delay, a shared memory space for storing the
image data was adopted. Consequently, there was a need for
an initialization function for allocating the memory space.
Receiving input parameters was also added to this function,
as it is not necessary to receive them separately for each
frame.