Multi-Band YCbCr Noise Modeling And Noise Reduction Based On Scene Metadata

摘要

Image enhancement is achieved by separating image signals, e.g. YCbCr image signals, into a series of frequency bands and performing noise reduction on bands below a given frequency but not on bands above that frequency. The bands are summed to develop the image enhanced signals. The YCbCr, multi-band approach to denoising is able to operate independently—and in an optimized fashion—on both luma and chroma channels. Noise reduction is done based on models developed for both luma and chroma channels by measurements taken for multiple frequency bands, in multiple patches on the ColorChecker chart, and at multiple gain levels, in order to develop a simple yet robust set of models that may be tuned off-line a single time for each camera and then applied to images taken by such cameras in real-time without excessive processing requirements and with satisfactory results across illuminant types and lighting conditions.

主权项

1. A non-transitory program storage device comprising instructions stored thereon to cause one or more processors to:
receive signals from an imager; process the signals received from the imager; separate at least one of the processed signals into a plurality of bands; apply noise reduction to all bands below a first frequency, wherein the instructions to apply comprise instructions to utilize an estimated noise level from a noise model, and wherein the instructions to apply further comprise instructions to evaluate one or more similarity thresholds for at least one pixel in the received signals based, at least in part, on one or more of the following: a band number of the respective pixel, an imager gain level when the signals were captured, and a gain-dependent desired denoising strength; sharpen all bands below the first frequency; and sum all of the bands to produce an enhanced image signal.