# US 2021/0241426 A1 · Method for Denoising Omnidirectional Videos and Rectified Videos > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0241426 A1 - Document ID: US20210241426A1 - Status: Published application - Kind: A1 - Application number: 16/613,932 - Filed: 2017-12-04 - Published: 2021-08-05 - Earliest claimed: 2017-12-04 - Assignees: Google Inc. - Applicants: Google LLC - Inventors: Damien Kelly; Neil Birkbeck; Balineedu Adsumilli; Mohammad Izadi - Web page: https://patentgrep.com/patent/US20210241426A1 ## CPC - G06T 5/70: Denoising; Smoothing - G06T 5/10: using non-spatial domain filtering - G06T 2207/10016: Video; Image sequence - G06T 2207/20021: Dividing image into blocks, subimages or windows - G06T 2207/20056: Discrete and fast Fourier transform, [DFT, FFT] - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract A method for denoising video content includes identifying a first frame block associated with a first frame of the video content. The method also includes estimating a first noise model that represents characteristics of the first frame block. The method also includes identifying at least one frame block adjacent to the first frame block. The method also includes generating a second noise model that represents characteristics of the at least one frame block adjacent to the first frame block by adjusting the first noise model based on at least one characteristic of the at least one frame block adjacent to the first frame block. The method also includes denoising the at least one frame block adjacent to the first frame block using the second noise model. ## Claims ### Claim 1 A method for denoising video content, the method comprising: identifying a first frame block associated with a first frame of the video content; estimating a first noise model that represents characteristics of the first frame block; identifying at least one frame block adjacent to the first frame block; generating a second noise model that represents characteristics of the at least one frame block adjacent to the first frame block by adjusting the first noise model based on at least one characteristic of the at least one frame block adjacent to the first frame block; and denoising the at least one frame block adjacent to the first frame block using the second noise model. ### Claim 2 (depends on 1) The method of claim 1, wherein the video content includes an omnidirectional video. ### Claim 3 (depends on 1) The method of claim 1, wherein the video content includes a 3-dimensional video. ### Claim 4 (depends on 1) The method of claim 1, wherein the video content includes a virtual reality video. ### Claim 5 (depends on 1) The method of claim 1, wherein estimating the first noise model includes estimating the first noise model based on at least a camera geometry. ### Claim 6 (depends on 5) The method of claim 5, wherein the camera geometry is estimated based on one or more characteristics of the video content. ### Claim 7 (depends on 1) The method of claim 1, wherein the characteristics of the first frame block include noise characteristics of the first frame block and the characteristics of the at least one frame block adjacent to the first frame block include noise characteristics of the at least one frame block adjacent to the first frame block. ### Claim 8 An apparatus for denoising video content, the apparatus comprising: a memory; and a processor, wherein the memory includes instructions executable by the processor to: identify a first frame block associated with a first frame of the video content; estimate a first noise model that represents characteristics of the first frame block; identify at least one frame block adjacent to the first frame block; generate a second noise model that represents characteristics of the at least one frame block adjacent to the first frame block by adjusting the first noise model based on at least one characteristic of the at least one frame block adjacent to the first frame block; and denoise the at least one frame block adjacent to the first frame block using the second noise model. ### Claim 9 (depends on 8) The apparatus of claim 8, wherein the video content includes an omnidirectional video. ### Claim 10 (depends on 8) The apparatus of claim 8, wherein the video content includes a 3-dimensional video. ### Claim 11 (depends on 8) The apparatus of claim 8, wherein the video content includes a virtual reality video. ### Claim 12 (depends on 8) The apparatus of claim 8, wherein the first noise model is estimated based on at least a camera geometry. ### Claim 13 (depends on 12) The apparatus of claim 12, wherein the camera geometry is estimated based on one or more characteristics of the video content. ### Claim 14 (depends on 8) The apparatus of claim 8, wherein the characteristics of the first frame block include noise characteristics of the first frame block and the characteristics of the at least one frame block adjacent to the first frame block include noise characteristics of the at least one frame block adjacent to the first frame block. ### Claim 15 An apparatus for denoising video content, the apparatus comprising: a memory; and a processor, wherein the memory includes instructions executable by the processor to: identify a first frame block associated with a first frame of the video content; estimate a first noise model that represents first noise power spectral density of the first frame block; identify at least one frame block adjacent to the first frame block; generate a second noise model that represents a second noise power spectral density of the at least one frame block adjacent to the first frame block by adjusting the first noise model based an estimated spatial variance between the first noise power spectral density and the second noise power spectral density; and denoise the at least one frame block adjacent to the first frame block using the second noise model. ### Claim 16 (depends on 15) The apparatus of claim 15, wherein the video content includes an omnidirectional video. ### Claim 17 (depends on 15) The apparatus of claim 15, wherein the video content includes a 3-dimensional video. ### Claim 18 (depends on 15) The apparatus of claim 15, wherein the video content includes a virtual reality video. ### Claim 19 (depends on 15) The apparatus of claim 15, wherein the first noise model is estimated based on at least a camera geometry. ### Claim 20 (depends on 15) The apparatus of claim 15, wherein the first noise power spectral density represents noise characteristics of the first frame block and the second noise power spectral density represents noise characteristics of the at least one frame block adjacent to the first frame block.