# US 12,602,741 B2 · Systems and methods regulating filter strength for temporal filtering > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,602,741 B2 - Document ID: US12602741B2 - Status: Granted - Kind: B2 - Application number: 18/465,851 - Filed: 2023-09-12 - Published: 2025-03-13 - Granted: 2026-04-14 - Earliest claimed: 2023-09-12 - Assignees: Microsoft Technology Licensing, LLC - Applicants: Microsoft Technology Licensing, LLC - Inventors: Michael Bleyer; Christian Markus Maekelae; Christopher Douglas Edmonds - Web page: https://patentgrep.com/patent/US12602741B2 ## CPC - G06T 5/10: using non-spatial domain filtering - G06T 5/50: using two or more images, e.g. averaging or subtraction - G06T 5/70: Denoising; Smoothing - G06T 7/248: {involving reference images or patches} - G06T 2207/20024: Filtering details - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract A system for regulating temporal filtering strength is configurable to: (i) obtain a light level indicator indicating light level associated with a real-world environment; (ii) determine a motion compensation confidence indicator using a current image and a previous image; (iii) determine a filter weight by processing the light level indicator and the motion compensation confidence indicator using a filter strength regulation module; and (iv) generate an output image by using at least the filter weight to filter a current frame with a previous frame. ## Claims ### Claim 1 A system for regulating filter strength for temporal filtering, comprising: one or more processors; and one or more computer-readable recording media that store instructions that are executable by the one or more processors to configure the system to: obtain a light level indicator indicating light level associated with a real-world environment; determine a motion compensation confidence indicator using a current image and a previous image; convert the motion compensation confidence indicator into a reliability value; determine a filter weight by processing the light level and the reliability value using a filter strength regulation module; and generate an output image by using at least the filter weight to filter a current frame with a previous frame. ### Claim 2 (depends on 1) The system of claim 1, wherein the light level indicator comprises a gain setting associated with the current frame. ### Claim 3 (depends on 1) The system of claim 1, wherein the previous image comprises a motion compensated previous image. ### Claim 4 (depends on 1) The system of claim 1, wherein determining the motion compensation confidence indicator comprises determining an optical flow magnitude between the current image and the previous image. ### Claim 5 (depends on 1) The system of claim 1, wherein processing the light level indicator and the reliability value using the filter strength regulation module comprises: determining a plurality of weights based on the reliability value and the light level indicator; and interpolating the plurality of weights to obtain the filter weight. ### Claim 6 (depends on 5) The system of claim 5, wherein the plurality of weights is selected using a weight function that, for each of a plurality of spatial noise settings, defines a weight for each of a plurality of light level indicator values. ### Claim 7 (depends on 6) The system of claim 6, wherein the weight function is generated by performing noise simulation to determine, for each particular spatial noise setting of the plurality of spatial noise settings, the weight for each of the plurality of light level indicator values that approximates, when applied to input imagery, the particular spatial noise setting. ### Claim 8 (depends on 6) The system of claim 6, wherein the weight for each particular light level indicator value of the plurality of light level indicator values is selected by: converting a target intensity value into a linear value; generating a plurality of images where each pixel comprises the linear value with simulated noise; for each potential weight value for the particular light level indicator value: generating a filtered image using the plurality of images as input; converting the filtered image into an intensity image; and determining noise characteristics of the intensity image; and defining the weight for the particular light level indicator value as the potential weight value that is associated with noise characteristics that approximate a target spatial noise setting. ### Claim 9 (depends on 1) The system of claim 1, wherein the previous frame comprises a previously generated output image. ### Claim 10 A system for regulating filter strength for temporal filtering, comprising: one or more image sensors; one or more processors; and one or more computer-readable recording media that store instructions that are executable by the one or more processors to configure the system to: capture an image frame of a real-world environment using the one or more image sensors; obtain a light level indicator indicating light level associated with the real-world environment; determine a filter weight by processing the light level indicator using a filter strength regulation module that implements a weight function generated via noise simulation, wherein the weight function defines a weight for each of a plurality of light level indicator values, wherein the weight for each particular light level indicator value of the plurality of light level indicator values is selected by: converting a target intensity value into a linear value; generating a plurality of images where each pixel comprises the linear value with simulated noise; for each potential weight value for the particular light level indicator value: generating a filtered image using the plurality of images as input; converting the filtered image into an intensity image; and determining noise characteristics of the intensity image; and defining the weight for the particular light level indicator value as the potential weight value that is associated with noise characteristics that approximate a target spatial noise setting; generate an output image by using at least the filter weight to filter a current frame with a previous frame; and present the output image on a display of the system. ### Claim 11 (depends on 10) The system of claim 10, wherein the light level indicator comprises a gain setting associated with the current frame. ### Claim 12 (depends on 10) The system of claim 10, wherein the previous frame comprises a previously generated output image. ### Claim 13 A system for regulating filter strength for temporal filtering, comprising: one or more processors; and one or more computer-readable recording media that store instructions that are executable by the one or more processors to configure the system to: determine a motion compensation confidence indicator using a current image and a previous image; convert the motion compensation confidence indicator into a reliability value; determine a filter weight by processing the reliability value using a filter strength regulation module; and generate an output image by using at least the filter weight to filter a current frame with a previous frame. ### Claim 14 (depends on 13) The system of claim 13, wherein the previous image comprises a motion compensated previous image. ### Claim 15 (depends on 13) The system of claim 13, wherein determining the motion compensation confidence indicator comprises determining an optical flow magnitude between the current image and the previous image. ### Claim 16 (depends on 13) The system of claim 13, wherein processing the reliability value using the filter strength regulation module comprises: determining a plurality of weights based on the reliability value; and interpolating the plurality of weights to obtain the filter weight. ### Claim 17 (depends on 1) The system of claim 1, wherein the reliability value is based on an optical flow mean magnitude associated with the motion compensation confidence indicator. ### Claim 18 (depends on 13) The system of claim 13, wherein the reliability value is based on an optical flow mean magnitude associated with the motion compensation confidence indicator.