# US 12,602,897 B2 · Face detection based filtering for image processing > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,602,897 B2 - Document ID: US12602897B2 - Status: Granted - Kind: B2 - Application number: 18/550,435 - Filed: 2021-05-21 - Published: 2024-05-09 - Granted: 2026-04-14 - Earliest claimed: 2021-05-21 - Assignees: QUALCOMM Incorporated - Applicants: QUALCOMM Incorporated - Inventors: Peiguang Ge; Changsheng Guo; Zhiqiang Jin - Web page: https://patentgrep.com/patent/US12602897B2 ## CPC - G06T 5/70: Denoising; Smoothing - G06V 10/25: Determination of region of interest [ROI] or a volume of interest [VOI] - G06T 5/20: using local operators - G06T 5/50: using two or more images, e.g. averaging or subtraction - G06V 10/761: {Proximity, similarity or dissimilarity measures} - G06V 40/161: {Detection; Localisation; Normalisation} - G06T 2207/10016: Video; Image sequence - G06T 2207/20221: Image fusion; Image merging - G06T 2207/30201: Face - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract An example device for image processing includes a memory and one or more processors coupled to the memory and configured to determine a region of interest (ROI) within a first frame and a second frame of a plurality of frames, apply a first amount of filtering on the ROI of the first frame and the second frame to generate a filtered ROI, apply a second amount of filtering on a remaining region of the first frame and the second frame to generate a filtered remaining region, combine the filtered ROI and the filtered remaining region to generate a filtered frame, and output the filtered frame. ## Claims ### Claim 1 A device for image processing, the device comprising: a memory; and one or more processors coupled to the memory and configured to: determine respective region of interests (ROIs) and respective remaining regions within a plurality of frames including a first frame and a second frame; apply a first amount of filtering on the ROIs to generate a filtered ROI, wherein to apply the first amount of filtering, the one or more processors are configured to determine a first group of frames from the plurality of frames to be included in filtering the ROIs based at least on a difference between an ROI in the first frame and an ROI in the second frame; apply a second amount of filtering on the remaining regions to generate a filtered remaining region, wherein to apply the second amount of filtering, the one or more processors are configured to determine a second group of frames from the plurality of frames to be included in filtering the remaining regions based at least on a difference between a remaining region in the first frame and a remaining region in the second frame; combine the filtered ROI and the filtered remaining region to generate a filtered frame; and output the filtered frame. ### Claim 2 (depends on 1) The device of claim 1, wherein the one or more processors are configured to detect a face in the first frame and the second frame, wherein the ROI in the first frame and the ROI in the second frame is a face ROI based on the detected face in the first frame and second frame. ### Claim 3 (depends on 1) The device of claim 1, wherein the one or more processors are configured to: determine that there is no ROI in a third frame; and apply the second amount of filtering on the third frame. ### Claim 4 (depends on 1) The device of claim 1, wherein the first group of frames includes a first number of frames, wherein the second group of frames includes a second number of frames, and wherein the first number of frames is less than the second number of frames. ### Claim 5 (depends on 1) The device of claim 1, wherein to determine the first group of frames, the one or more processors are configured to determine the first group of frames based at least on a comparison of the difference between the ROI in the first frame and the ROI in the second frame to a first threshold, wherein to determine the second group of frames, the one or more processors are configured to determine the second group of frames based at least on a comparison of the difference between the remaining region in the first frame and the remaining region in the second frame to a second threshold, and wherein the first threshold and the second threshold are different. ### Claim 6 (depends on 1) The device of claim 1, wherein to apply the first amount of filtering on the ROIs, the one or more processors are configured to: determine a first difference value between the ROI in the first frame and the ROI in the second frame; compare the first difference value to a first threshold; and blend the ROI in the first frame with the ROI in the second frame based on the comparison of the first difference value to the first threshold to generate an intermediate filtered ROI prior to generating the filtered ROI, and wherein to apply the second amount of filtering on the remaining regions, the one or more processors are configured to: determine a second difference value between the remaining region in the first frame and the remaining region in the second frame; compare the second difference value to a second threshold, wherein the second threshold is different than the first threshold; and blend the remaining region in the first frame with the remaining region in the second frame based on the comparison of the second difference value to the second threshold to generate an intermediate filtered remaining region prior to generating the filtered remaining region. ### Claim 7 (depends on 6) The device of claim 6, wherein to apply the first amount of filtering on the ROIs, the one or more processors are configured to: determine a third difference value between the intermediate filtered ROI and the ROI in a third frame; compare the third difference value to the first threshold; and blend the intermediate filtered ROI with the ROI in the third frame based on the comparison of the third difference value to the first threshold to generate the filtered ROI, and wherein to apply the second amount of filtering on the remaining regions, the one or more processors are configured to: determine a fourth difference value between the intermediate filtered remaining region and the remaining region in the third frame; compare the fourth difference value to the second threshold; and blend the intermediate filtered remaining region with the remaining region in the third frame based on the comparison of the fourth difference value to the second threshold to generate the filtered remaining region. ### Claim 8 (depends on 6) The device of claim 6, wherein to blend the ROI in the first frame with the ROI in the second frame based on the comparison of the first difference value to the first threshold, the one or more processors are configured to: determine that the first difference value is less than the first threshold; and blend the ROI in the first frame with the ROI in the second frame based on the first difference value being less than the first threshold, wherein to blend the remaining region in the first frame with the remaining region in the second frame based on the comparison of the second difference value to the second threshold, the one or more processors are configured to: determine that the second difference value is less than the second threshold; and blend the remaining region in the first frame with the remaining region in the second frame based on the second difference value being less than the second threshold, and wherein the first threshold is less than the second threshold. ### Claim 9 (depends on 1) The device of claim 1, wherein the plurality of frames are captured with a camera within less than 300 milliseconds and capture substantially same image content. ### Claim 10 (depends on 1) The device of claim 1, wherein the device comprises a mobile device that includes one or more cameras to capture the plurality of frames. ### Claim 11 A method for image processing, the method comprising: determining respective region of interests (ROIs) and respective remaining regions within a plurality of frames including a first frame and a second frame; applying a first amount of filtering on the ROIs to generate a filtered ROI, wherein applying the first amount of filtering comprises determining a first group of frames from the plurality of frames to be included in filtering the ROIs based at least on a difference between an ROI in the first frame and an ROI in the second frame; applying a second amount of filtering on the remaining regions to generate a filtered remaining region, wherein applying the second amount of filtering comprises determining a second group of frames from the plurality of frames to be included in filtering the remaining regions based at least on a difference between a remaining region in the first frame and a remaining region in the second frame; combining the filtered ROI and the filtered remaining region to generate a filtered frame; and outputting the filtered frame. ### Claim 12 (depends on 11) The method of claim 11, further comprising detecting a face in the first frame and the second frame, wherein the ROI in the first frame and the ROI in the second frame is a face ROI based on the detected face in the first frame and the second frame. ### Claim 13 (depends on 11) The method of claim 11, further comprising: determining that there is no ROI in a third frame; and applying the second amount of filtering on the third frame. ### Claim 14 (depends on 11) The method of claim 11, wherein the first group of frames includes a first number of frames, wherein the second group of frames includes a second number of frames, and wherein the first number of frames is less than the second number of frames. ### Claim 15 (depends on 11) The method of claim 11, wherein determining the first group of frames comprises determining the first group of frames based at least on a comparison of the difference between the ROI in the first frame and the ROI in the second frame to a first threshold, wherein determining the second group of frames comprises determining the second group of frames based at least on a comparison of the difference between the remaining region in the first frame and the remaining region in the second frame to a second threshold, and wherein the first threshold and the second threshold are different. ### Claim 16 (depends on 11) The method of claim 11, wherein applying the first amount of filtering on the ROIs comprises: determining a first difference value between the ROI in the first frame and the ROI in the second frame; comparing the first difference value to a first threshold; and blending the ROI in the first frame with the ROI in the second frame based on the comparison of the first difference value to the first threshold to generate an intermediate filtered ROI prior to generating the filtered ROI, and wherein applying the second amount of filtering on the remaining regions comprises: determining a second difference value between the remaining region in the first frame and the remaining region in the second frame; comparing the second difference value to a second threshold, wherein the second threshold is different than the first threshold; and blending the remaining region in the first frame with the remaining region in the second frame based on the comparison of the second difference value to the second threshold to generate an intermediate filtered remaining region prior to generating the filtered remaining region. ### Claim 17 (depends on 16) The method of claim 16, wherein applying the first amount of filtering on the ROIs comprises: determining a third difference value between the intermediate filtered ROI and the ROI in a third frame; comparing the third difference value to the first threshold; and blending the intermediate filtered ROI with the ROI in the third frame based on the comparison of the third difference value to the first threshold to generate the filtered ROI, and wherein applying the second amount of filtering on the remaining regions comprises: determining a fourth difference value between the intermediate filtered remaining region and the remaining region in the third frame; comparing the fourth difference value to the second threshold; and blending the intermediate filtered remaining region with the remaining region in the third frame based on the comparison of the fourth difference value to the second threshold to generate the filtered remaining region. ### Claim 18 (depends on 16) The method of claim 16, wherein blending the ROI in the first frame with the ROI in the second frame based on the comparison of the first difference value to the first threshold comprises: determining that the first difference value is less than the first threshold; and blending the ROI in the first frame with the ROI in the second frame based on the first difference value being less than the first threshold, wherein blending the remaining region in the first frame with the remaining region in the second frame based on the comparison of the second difference value to the second threshold comprises: determining that the second difference value is less than the second threshold; and blending the remaining region in the first frame with the remaining region in the second frame based on the second difference value being less than the second threshold, and wherein the first threshold is less than the second threshold. ### Claim 19 (depends on 11) The method of claim 11, further comprising capturing the plurality of frames with a camera within less than 300 milliseconds that capture substantially same image content. ### Claim 20 A computer-readable storage medium storing instructions thereon that when executed cause one or more processors to: determine respective region of interests (ROIs) and respective remaining region within a plurality of frames including a first frame and a second frame; apply a first amount of filtering on the ROIs to generate a filtered ROI, wherein the instructions that cause the one or more processors to apply the first amount of filtering comprise instructions that cause the one or more processors to determine a first group of frame from the plurality of frames to be included in filtering the ROIs based at least on a difference between an ROI in the first frame and an ROI in the second frame; apply a second amount of filtering on the remaining regions to generate a filtered remaining region, wherein the instructions that cause the one or more processors to apply the second amount of filtering comprise instructions that cause the one or more processors to determine a second group of frames from the plurality of frames to be included in filtering the remaining regions based at least on a difference between a remaining region in the first frame and a remaining region in the second frame; combine the filtered ROI and the filtered remaining region to generate a filtered frame; and output the filtered frame. ### Claim 21 (depends on 20) The computer-readable storage medium of claim 20, wherein the instructions further comprise instructions that cause the one or more processors to: determine that there is no ROI in a third frame; and apply the second amount of filtering on the third frame. ### Claim 22 (depends on 20) The computer-readable storage medium of claim 20, wherein the instructions that cause the one or more processors to determine the first group of frames comprise instructions that cause the one or more processors to determine the first group of frames based at least on a comparison of the difference between the ROI in the first frame and the ROI in the second frame to a first threshold, wherein the instructions that cause the one or more processors to determine the second group of frame comprise instructions that cause the one or more processors to determine the second group of frames based at least on a comparison of the difference between the remaining region in the first frame and the remaining region in the second frame to a second threshold, and wherein the first threshold and the second threshold are different. ### Claim 23 A device for image processing, the device comprising: means for determining respective region of interests (ROIs) and respective remaining portions within a plurality of frames including a first frame and a second frame; means for applying a first amount of filtering on the ROIs to generate a filtered ROI, wherein the means for applying the first amount of filtering comprises means for determining a first group of frames from the plurality of frames to be included in filtering the ROIs based at least on a difference between an ROI in the first frame and an ROI in the second frame; means for applying a second amount of filtering on the remaining regions to generate a filtered remaining region, wherein the means for applying the second amount of filtering comprises means for determining a second group of frames from the plurality of frames to be included in filtering the remaining regions based at least on a difference between a remaining region in the first frame and a remaining region in the second frame; means for combining the filtered ROI and the filtered remaining region to generate a filtered frame; and means for outputting the filtered frame. ### Claim 24 (depends on 23) The device of claim 23, further comprising means for detecting a face in the first frame and the second frame, wherein the ROI in the first frame and the ROI in the second frame is a face ROI based on the detected face in the first frame and the second frame.