IP Library Granted Patent US 12705810
Granted Patent B2
US 12705810 · App. 18/608,381 · Granted Aug 11, 2026

Real time film grain rendering and parameter estimation

Inventors: Daizong Tian (Evanston, IL); Jingxian Wang (Evanston, IL); Kaixuan Zhang (Evanston, IL); Thrasyvoulos N. Pappas (Evanston, IL)
Assignee: Northwestern University
G06T11/60
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12705810
App. No.
18/608,381
Granted
Aug 11, 2026
Kind
B2
Abstract

A method for film grain rendering includes receiving, by a computing system a first image. The method also includes modeling, by a processor of the computing system, film grains for the image. The method also includes deriving, by the processor, second order statistics for the film grains. The method also includes synthesizing, by the processor and based at least in part on the second order statistics, the film grains as correlated signal dependent noise. The method further includes rendering, by the processor, a second image that corresponds to the first image and that includes the synthesized film grains.

Claims (32)

1 . A method for film grain rendering, the method comprising:

receiving, by a computing system, a first image;

modeling, by a processor of the computing system, film grains for the first image;

identifying, by the processor, one or more portions of the first image that have a constant value;

deriving, by the processor, second order statistics for the film grains, wherein deriving the second order statistics is based on the one or more portions of the first image that have the constant value;

synthesizing, by the processor and based at least in part on the second order statistics, the film grains as correlated signal dependent noise; and

rendering, by the processor, a second image that corresponds to the first image and that includes the synthesized film grains.

2 . The method of claim 1 , wherein the second order statistics for the film grains include variance.

3 . The method of claim 1 , wherein the second-order statistics for the film grains include correlation.

4 . The method of claim 1 , wherein the deriving is performed based on an assumption that the first image has a constant value.

5 . The method of claim 1 , further comprising deriving, by the processor, one or more formulas for film grain parameter estimation based at least in part on the derived second order statistics.

6 . The method of claim 5 , wherein the one or more formulas for film grain parameter estimation are based on an assumption of constant film grain radius.

7 . The method of claim 1 , wherein the film grains are modeled as a filtered Boolean model.

8 . The method of claim 1 , wherein the synthesizing further comprises uniformly distributing the film grains throughout the first image.

9 . The method of claim 8 , wherein each of the film grains includes a grain center, and wherein the grain centers are distributed in accordance with a Poisson process that is based at least in part on a parameter that controls grain density.

10 . The method of claim 9 , wherein each of the film grains includes a radius, and wherein the grain radii are identically and independently distributed throughout the first image.

11 . The method of claim 10 , wherein the radii are identically and independently distributed throughout the first image in accordance with a density function, and wherein the radii are distributed independent of the grain centers.

12 . A system for rendering film grain, the system comprising:

a memory that stores a first image; and

a processor operatively coupled to the memory, wherein the processor is configured to:

model film grains for the first image;

identify one or more portions of the first image that have a constant value;

derive second order statistics for the film grains, wherein the second order statistics are derived based on the one or more portions of the first image that have the constant value;

synthesize, based at least in part on the second order statistics, the film grains as correlated signal dependent noise; and

render a second image that corresponds to the first image and that includes the synthesized film grains.

13 . The system of claim 12 , wherein the second order statistics for the film grains include variance and correlation.

14 . The system of claim 12 , wherein the processor is further configured to derive one or more formulas for film grain parameter estimation based at least in part on the derived second order statistics.

15 . The system of claim 14 , wherein the one or more formulas for film grain parameter estimation are based on an assumption of constant film grain radius.

16 . The system of claim 12 , wherein the processor models the film grains as a filtered Boolean model.

17 . The system of claim 12 , wherein each of the film grains includes a grain center, and wherein the processor distributes the grain centers in accordance with a Poisson process.

18 . The system of claim 17 , wherein each of the film grains includes a radius, and wherein the processor identically and independently distributes the radii throughout the first image.

19 . The system of claim 18 , wherein the radii are distributed throughout the first image in accordance with a density function, and wherein the radii are distributed independent of the grain centers.