Video and Vision Processing Suite Intel® FPGA IP User Guide

ID 683329
Date 4/04/2022
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

15.1. About the Guard Bands IP

The IP compares each color plane in the input video to upper and lower guard band values. If the value in any color plane exceeds the upper guard band, the IP replaces the value by the upper guard band. Similarly, if the value in any color plane falls below the lower guard band, the IP replaces the value by the lower guard band.

You can specify different guard band values for each color plane. You can alter these values at run-time through the Avalon memory-mapped interface. Otherwise, the guard band values are fixed at compile-time.

The input can be unsigned data or signed 2’s complement data for which the IP converts the data to an unsigned format (by adding half the maximum range) before applying guard bands. The guard bands are specified as unsigned values.

The IP can drive the output data as signed data but, as with signed input data, the IP applies guard bands on the unsigned data before it converts it to signed output. The IP converts the output data to a signed format by subtracting half the maximum range after applying guard bands.

You cannot select both signed input and signed output data. The IP prepares data for other video and vision processing IPs, which primarily operate on unsigned data. You can place this IP before and after another video and vision IP, configuring the first for signed input and the second to use signed output data.