The FDIF module performs face detection within a picture stored in memory (QVGA luminance data
resolution). This module is typically used for video encoding, face-based priority auto-focusing, or red-eye
removal.
The face detect (FD) core is a stand-alone module: it embeds its own direct memory access (DMA) engine
for accessing data in memory. The FD core is under microprocessor unit (MPU) control for its initialization
and to start the processing operation.
The FD core supports single input resolution (QVGA) in a single format (8-bit Luma). The FD core requires
that the input image is stored in synchronous dynamic random access memory (SDRAM). The core also
requires working memory mapped in SDRAM. The input image data (320 Ã 240) requires 75KiB for the
SDRAM and 51.25KiB for the working memory.