Denoising binned-depth images

Number: 10565685
Published: 2020-02-18
Filed: 2017-06-27
Assignee: Disney Enterprises, Inc.
Inventors: Adler; David M., Vicini; Delio Aleardo, Burley; Brent, Novak; Jan, Rousselle; Fabrice Pierre Armand
CPC: G06T5/70
Verdict: Low Notable software
Source: Google Patents · FreePatentsOnline

The keeper's note

Depth-image denoising technique.

Abstract

According to one implementation, an image rendering system includes a computing platform having a hardware processor and a system memory storing an image denoising software code. The hardware processor executes the image denoising software code to receive an image file including multiple pixels, each pixel containing multiple depth bins, and to select a pixel including a noisy depth bin from among the pixels for denoising. The hardware processor further executes the image denoising software code to identify a plurality of reference depth bins from among the depth bins contained in one or more of the pixels, for use in denoising the noisy depth bin, and to denoise the noisy depth bin using an average of depth bin values corresponding respectively to each of the reference depth bins.

Background

BACKGROUND(1) Deep compositing is an image rendering technique for generating complex images from multiple sources. Unlike “flat” images used in traditional compositing, in which each pixel includes a single pixel value, each deep pixel of a deep image can contain multiple pixel values representing contributions from objects at different camera depths. Having multiple depth bin values available to the compositor advantageously enables capabilities such as inserting atmosphere, or isolating color corrections and lighting adjustments to particular depths.(2) Unfortunately, however, rendered images, flat images as well as deep images, are often plagued with residual noise produced during their rendering. Although sophisticated sampling strategies can reduce the noise, avoiding it completely inside the renderer is nearly impossible. Moreover, although post rendering denoising techniques capable of producing high-quality outputs for flat images exist, none of those existing techniques can preserve depth decompositions as required for the denoising of deep images.SUMMARY(3) There are provided systems and methods for denoising binned-depth images, substantially as shown in and/or described in connection with at least one of the figures, and as set forth more completely in the claims.

Claims

1. An image rendering system comprising: a computing platform including a hardware processor and a system memory; an image denoising software code stored in the system memory; the hardware processor configured to execute the image denoising software code to: receive an image file including a plurality of pixels, each of the plurality of pixels containing multiple depth bins each depth bin representing a different depth and having a corresponding color data; select a pixel including a noisy depth bin from among the plurality of pixels; identify a plurality of reference depth bins from among the multiple depth bins contained in the plurality of pixels, for use in denoising the noisy depth bin; and denoise the noisy depth bin using an average of depth bin values corresponding respectively to each of the plurality of reference depth bins. 8. A method for use by an image rendering system including a computing platform having a hardware processor and a system memory storing an image denoising software code, the method comprising: receiving, using the hardware processor, an image file including a plurality of pixels, each of the plurality of pixels containing multiple depth bins each depth bin representing a different depth and having a corresponding color data; selecting, using the hardware processor, a pixel including a noisy depth bin from among the plurality of pixels; identifying, using the hardware processor, a plurality of reference depth bins from among the multiple depth bins contained in the plurality of pixels, for use in denoising the noisy depth bin; and denoising, using the hardware processor, the noisy depth bin using an average of depth bin values corresponding respectively to each of the plurality of reference depth bins. 15. A computer-readable non-transitory medium having stored thereon instructions, which when executed by a hardware processor, instantiate a method comprising: receiving an image file including a plurality of pixels, each of the plurality of pixels containing multiple depth bins each depth bin representing a different depth and having a corresponding color data; selecting a pixel including a noisy depth bin from among the plurality of pixels; identifying a plurality of reference depth bins from among the multiple depth bins contained in the plurality of pixels, for use in denoising the noisy depth bin; and denoising the noisy depth bin using an average of depth bin values corresponding respectively to each of the plurality of reference depth bins.