# US 11,113,789 B2 · Method and device for digital image restoration > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,113,789 B2 - Document ID: US11113789B2 - Status: Granted - Kind: B2 - Application number: 16/347,518 - Filed: 2017-11-03 - Published: 2019-09-19 - Granted: 2021-09-07 - Earliest claimed: 2016-11-04 - Assignees: Centre National D'Etudes Spatiales - Applicants: Centre National d'Études Spatiales - Inventors: Gwendoline Blanchet; Jean-Marc Delvit; Christophe Latry; Carole Thiebaut; Roberto Camarero - Web page: https://patentgrep.com/patent/US11113789B2 ## CPC - G06T 5/70: Denoising; Smoothing - G06T 5/10: using non-spatial domain filtering - G06T 5/50: using two or more images, e.g. averaging or subtraction - G06T 9/00: Image coding (bandwidth or redundancy reduction for static pictures H04N1/41; coding or decoding of static colour picture signals H04N1/64; methods or arrangements for coding, decoding, compressing or decompressing digital video signals H04N19/00) - G06T 2207/10036: Multispectral image; Hyperspectral image - G06T 2207/20182: Noise reduction or smoothing in the temporal domain; Spatio-temporal filtering - G06T 2207/20221: Image fusion; Image merging - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL ## Abstract The invention concerns a method and a device for digital image restoration. The digital image to be restored comes from an initial image acquired by an image acquisition device having an associated acquisition instrumental noise. The method comprises the following steps: —obtaining (30) an intermediate digital image with restoration of the acquisition instrumental noise, —denoising (32) the intermediate digital image in order to obtain a denoised intermediate digital image, —deconvolution (34) of the denoised digital image in order to obtain a restored digital image. ## Claims ### Claim 1 A method for restoring a digital image, the digital image to be restored being derived from an initial image acquired by an image acquisition device having an associated acquisition instrumental noise, the method including the steps of: obtaining an intermediate digital image with restitution of the associated acquisition instrumental noise; denoising the intermediate digital image to obtain a denoised intermediate digital image; and deconvoluting the denoised intermediate digital image to obtain a restored digital image, wherein the step of obtaining an intermediate digital image comprises sub-steps of: obtaining, by applying a transformation called compression transformation, of a representation of the digital image to be restored by a plurality of coefficient blocks, each coefficients block of the plurality of coefficient blocks corresponding to a pixels block of the digital image to be restored; for at least one processed coefficient of a coefficients block: calculating a noise threshold as a function of a value representative of an image acquisition instrumental noise model; comparing an absolute value of the at least one processed coefficient to the noise threshold; and when the absolute value of the at least one processed coefficient is higher than or equal to the noise threshold, the at least one processed coefficient is left unchanged, when the absolute value of the at least one processed coefficient is lower than the noise threshold, replacing the absolute value of the at least one processed coefficient with a noise value depending on the value representative of the image acquisition noise model. ### Claim 2 A method for restoring a digital image, the digital image to be restored being derived from an initial image acquired by an image acquisition device having an associated acquisition instrumental noise, the method including the steps of: obtaining an intermediate digital image with restitution of the associated acquisition instrumental noise; denoising the intermediate digital image to obtain a denoised intermediate digital image; deconvoluting the denoised intermediate digital image to obtain a restored digital image; wherein the associated acquisition instrumental noise is modeled by a model parameterized by two coefficients, the two coefficients having values used during the restitution of the associated acquisition instrumental noise, and wherein the denoising step includes the application of a stabilization transformation of a parameterized variance by the two coefficients characterizing a parameterized model of the associated acquisition instrumental noise to provide a variance stabilization; and wherein the denoising step includes the following sub-steps of: applying the variance stabilization transformation to obtain a stabilized intermediate image; applying a stationary noise denoising method on the stabilized intermediate image; and applying an inverse transformation of the variance stabilization transformation on the digital image resulting from the step of denoising the intermediate digital image to obtain a denoised intermediate digital image. ### Claim 3 (depends on 1) The method for restoring a digital image according to claim 1, wherein the digital image to be restored has been obtained by applying a compression to the initial image acquired by the image acquisition device to obtain a compressed initial image, and then a decompression of the compressed initial image. ### Claim 4 (depends on 3) The method for restoring a digital image according to claim 3, wherein the restitution of the associated acquisition instrumental noise is performed in connection with the decompression of the compressed initial image. ### Claim 5 (depends on 1) The method for restoring a digital image according to claim 1, wherein the digital image to be restored is a multispectral image composed of a plurality of acquired images in different spectral bands and wherein the steps of obtaining an intermediate digital image and of denoising the intermediate digital image are applied to each of the plurality of acquired images in different spectral bands, and the deconvolution step is applied to at least one intermediate digital images obtained from the plurality of acquired images. ### Claim 6 (depends on 5) The method for restoring a digital image according to claim 5, further including an image fusion step applied after denoising or after deconvolution, to obtain a final restored image. ### Claim 7 (depends on 2) The method for restoring a digital image according to claim 2, wherein the digital image to be restored is a multispectral image composed of a plurality of acquired images in different spectral bands, and wherein the values of the two coefficients of the model depend on an acquisition spectral band and are determined for each acquisition spectral band. ### Claim 8 (depends on 2) The method for restoring a digital image according to claim 2, wherein the variance stabilization transformation is an Anscombe transformation. ### Claim 9 (depends on 1) The method for restoring a digital image according to claim 1, wherein the associated acquisition instrumental noise is modeled by a parameterized model defining a noise standard deviation σ by σ=√{square root over (a2+b·s)}, where a and b are coefficients of the parameterized model and s is a coefficient value representative of an average of values associated with pixels of a pixels block of the digital image. ### Claim 10 (depends on 1) A computer program product, comprising: a non-transitory computer readable storage medium having program instructions for implementing the steps of a method for restoring a digital image according to claim 1 during execution of the computer program product by a processor of a programmable device. ### Claim 11 (depends on 3) The method for restoring a digital image according to claim 3, wherein the digital image to be restored is a multispectral image composed of a plurality of acquired images in different spectral bands and wherein the steps of obtaining an intermediate digital image and of denoising the intermediate digital image are applied to each of the plurality of acquired images in different spectral bands, and the deconvolution step is applied to at least one intermediate digital image obtained from the plurality of acquired images. ### Claim 12 (depends on 4) The method for restoring a digital image according to claim 4, wherein the digital image to be restored is a multispectral image composed of a plurality of acquired images in different spectral bands and wherein the steps of obtaining an intermediate digital image and of denoising the intermediate digital image are applied to each of the plurality of acquired images in different spectral bands, and the deconvolution step is applied to at least one intermediate digital image obtained from the plurality of acquired images. ### Claim 13 (depends on 12) The method for restoring a digital image according to claim 12, further including an image fusion step applied after denoising or after deconvolution, to obtain a final restored image. ### Claim 14 (depends on 13) The method for restoring a digital image according to claim 13, wherein the acquisition instrumental noise is modeled by a model parameterized by two coefficients, the two coefficients having values used during the restitution of the associated acquisition instrumental noise, and wherein the denoising step includes the application of a stabilization transformation of a parameterized variance by the two coefficients characterizing a parameterized model of the associated acquisition instrumental noise. ### Claim 15 (depends on 14) The method for restoring a digital image according to claim 14, wherein the digital image to be restored is a multispectral image composed of a plurality of acquired images in different spectral bands, and wherein values of the two coefficients depend on the different spectral bands and are determined for each different spectral band. ### Claim 16 (depends on 15) The method for restoring a digital image according to claim 15, wherein the denoising step includes the following sub-steps of: applying the variance stabilization transformation to obtain a stabilized intermediate image, applying a stationary noise denoising method on the stabilized intermediate image, applying an inverse transformation of the variance stabilization transformation on a denoised intermediate digital image resulting from the step of applying a denoising method to obtain the denoised intermediate digital image.