The Vision Pre-processing Accelerator (VPAC) is a set of common vision primitive functions, performing memory-to-memory (M2M) pixel data processing tasks, such as: color processing and enhancement, noise filtering, wide dynamic range (WDR) processing, lens distortion correction, pixel remap for dewarping, on-the-fly scale generation, on-the-fly pyramid generation. The VPAC offloads these common tasks from the main SoC processors (ARM, DSP, etc.), so these CPUs can be utilized for differentiated high-level algorithms. The VPAC is designed to support multiple cameras by working in time-multiplexing mode. The VPAC works as front end to vision processing and prepares frame/scales for further processing by other vision accelerators or processor cores in the SoC.

For more information, see Vision Pre-processing Accelerator (VPAC) section in Processors and Accelerators chapter in the device TRM.