# US 12,762,659 B2 · Image processing apparatus for performing noise reduction processing > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,762,659 B2 - Document ID: US12762659B2 - Status: Granted - Kind: B2 - Application number: 18/058,099 - Filed: 2022-11-22 - Published: 2023-05-25 - Granted: 2026-10-13 - Earliest claimed: 2021-11-25 - Assignees: Canon Kabushiki Kaisha - Applicants: Canon Kabushiki Kaisha - Inventors: Tsuyoshi Kobayashi - Web page: https://patentgrep.com/patent/US12762659B2 ## CPC - G06T 5/70: Denoising; Smoothing - G06T 5/50: using two or more images, e.g. averaging or subtraction - G06T 5/73: Deblurring; Sharpening - G06T 2207/10121: Fluoroscopy - G06T 2207/20081: Training; Learning - G06T 2207/20216: Image averaging - G06T 2207/30148: Semiconductor; IC; Wafer - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract An image processing apparatus includes an obtaining unit configured to obtain a first radiation image of an inspection target object that has been captured by a radiation imaging apparatus, and a generation unit configured to generate a second radiation image including noise reduced as compared with the first radiation image, by inputting the first radiation image obtained by the obtaining unit, to a learned model obtained by performing learning using learning data including a radiation image to which noise simulating system noise of a radiation imaging apparatus in accordance with a distribution that is based on a manufacturing variation of a radiation imaging apparatus is added. ## Claims ### Claim 1 An image processing apparatus comprising: at least one processor; and a memory coupled to the at least one processor, the memory having instructions that, when executed by the at least one processor, cause the at least one processor to function as: an obtaining unit configured to obtain a first radiation image of an inspection target object that has been captured by a radiation imaging apparatus; and a generation unit configured to generate a second radiation image including noise reduced as compared with the first radiation image, by inputting the first radiation image obtained by the obtaining unit to a learned model obtained by performing learning using learning data including a radiation image to which noise simulating system noise of the radiation imaging apparatus and noise simulating a quantum noise included in radiation that have been obtained by using a distribution that is based on a manufacturing variation among radiation imaging apparatuses are added, wherein the noise simulating system noise is obtained by using a distribution indicating a variation among a plurality of system noise characteristics, that is generated by the manufacturing variation among the radiation imaging apparatuses, and being obtained by using an average and a variance of the plurality of system noise characteristics. ### Claim 2 (depends on 1) The image processing apparatus according to claim 1, wherein the generation unit obtains the noise simulating system noise by using a system noise characteristic selected at random from among the plurality of system noise characteristics. ### Claim 3 (depends on 1) The image processing apparatus according to claim 1, wherein the generation unit obtains the noise simulating the quantum noise by using a distribution indicating an in-plane characteristic variation of a modulation transfer function (MTF) of a fluorescent member arranged within the radiation imaging apparatus, the in-plane characteristic variation being generated based on the manufacturing variation among the radiation imaging apparatuses. ### Claim 4 (depends on 3) The image processing apparatus according to claim 3, wherein the generation unit obtains, as the noise simulating the quantum noise, noise in which a high-frequency component included in the quantum noise has been attenuated by using the distribution indicating the in-plane characteristic variation of the MTF. ### Claim 5 (depends on 3) The image processing apparatus according to claim 3, wherein the generation unit obtains the distribution indicating the in-plane characteristic variation of the MTF by using an average and a variance of a plurality of characteristics of the MTF. ### Claim 6 (depends on 5) The image processing apparatus according to claim 5, wherein the generation unit obtains the noise simulating the quantum noise by using a characteristic of the MTF selected at random from among the plurality of characteristics of the MTF. ### Claim 7 (depends on 1) The image processing apparatus according to claim 1, wherein the radiation imaging apparatuses are a plurality of manufactured radiation imaging apparatuses of one model, and wherein the generation unit obtains the distribution that is based on the manufacturing variation among the radiation imaging apparatuses by using statistical information obtained based on a manufacturing record of the plurality of manufactured radiation imaging apparatuses of the one model. ### Claim 8 (depends on 1) The image processing apparatus according to claim 1, wherein the radiation imaging apparatuses are a plurality of manufactured radiation imaging apparatuses of different models, and wherein the generation unit obtains the distribution that is based on the manufacturing variation among the radiation imaging apparatuses by using statistical information obtained based on a manufacturing record of the plurality of manufactured radiation imaging apparatuses of the different models. ### Claim 9 An image processing method comprising: obtaining a first radiation image of an inspection target object that has been captured by a radiation imaging apparatus; and generating a second radiation image including noise reduced as compared with the first radiation image, by inputting the obtained first radiation image to a learned model obtained by performing learning using learning data including a radiation image to which noise simulating system noise of the radiation imaging apparatus and noise simulating a quantum noise included in radiation that have been obtained by using a distribution that is based on a manufacturing variation among radiation imaging apparatuses are added, wherein the noise simulating the system noise is obtained by using a distribution indicating a variation among a plurality of system noise characteristics, that is generated by the manufacturing variation among the radiation imaging apparatuses, and being obtained by using an average and a variance of the plurality of system noise characteristics. ### Claim 10 (depends on 9) A non-transitory computer-readable medium storing a program for causing a computer to execute the image processing method according to claim 9.