The development of this thesis achieved the proposed goals: it was developed an hardware module capable of performing real time color space conversion of video streams. This module supports all the color spaces presented in the standards proposed, performs downsampling and upsampling to Y’Cr’Cb 4:2:2 formats with two different FIR filter implementations and supports pixel repetition and 3D video structures. It was verified by comparing with a software model implementing its expected behavior and it was successfully synthesized to a 40nm technology, achieving the goal frequency of 600 MHz.

The module developed is innovative in the sense that the market and bibliographic research didn’t provide results of other cores that implement the full color space conversion chain, from any RGB/YCrCb color space to any RGB/YCrCb color space. The majority of the available solutions only consider RGB-YCrCb conversion.

Therefore, this module offers a real-time hardware accelerated solution for a task very computationally intense, which effort is growing with the increase of video frame resolutions, following the industry tendency of moving imaging computational exigent tasks to dedicated hardware with the arising of embedded image and vision processing solutions from several IP vendors.