Abstract: To generate a favorable image that suppresses the occurrence of pattern noise in a method of performing patch-based noise reduction for a RAW image. In order to attain this object, the image processing apparatus of the present invention includes a pixel setting unit; a patch setting unit; and a noise reduction unit. Then, the pixel setting unit sets, for one pixel of interest, at least two kinds of pixel whose positions are different in a minimum unit of the color filter array as the plurality of reference pixels. Further, the patch setting unit sets, for one pixel of interest, at least two kinds of reference patch whose shapes are different.

Abstract: An image processing device that corrects an image captured by an imaging sensor capable of setting an exposure condition for each of a plurality of regions includes an acquisition unit configured to acquire the image by setting the exposure condition for each region in the imaging sensor, derivation unit configured to, regarding a boundary between a region of interest and at least one adjacent region adjacent to the region of interest in the image, derive a correction amount in the region of interest based on pixel values of a boundary pixel group in contact with the boundary included in the region of interest and pixel values of a boundary pixel group in contact with the boundary included in the adjacent region, and correction unit configured to correct pixel values of pixels in the region of interest based on the derived correction amount.

Abstract: A non-local means method is insufficient in its noise reduction effect or edge retainability due to a perfect match between blocks in a case where a reference pixel matches a target pixel. Therefore, information on a target region and plural reference regions is obtained for the target pixel. Whether the target region matches any one of the reference regions is determined from the obtained information. Switching between weight derivation methods based on similarity between the target region and the reference region is performed according to a determined result.

Abstract: To generate a favorable image that suppresses the occurrence of pattern noise in a method of performing patch-based noise reduction for a RAW image. In order to attain this object, the image processing apparatus of the present invention includes a pixel setting unit; a patch setting unit; and a noise reduction unit. Then, the pixel setting unit sets, for one pixel of interest, at least two kinds of pixel whose positions are different in a minimum unit of the color filter array as the plurality of reference pixels. Further, the patch setting unit sets, for one pixel of interest, at least two kinds of reference patch whose shapes are different.

Abstract: An image processing apparatus includes a processing unit to perform noise reduction processing on image data and an output unit to output noise-reduced output image data. The processing unit determines a weight for a reference pixel based on a target area including a target pixel and a reference area including the reference pixel and sets, as a pixel value of the target pixel on which processed noise reduction processing has been executed, a value calculated using reference pixels and a weight corresponding to each of the reference pixels. The size of the target area equals to the size of the reference area. The output unit outputs, as the pixel value of the target pixel after noise reduction processing, either a first value, calculated when a first pixel group is set as the target area, or a second value, calculated when a second pixel group is set as the target area.

Abstract: A non-local means method is insufficient in its noise reduction effect or edge retainability due to a perfect match between blocks in a case where a reference pixel matches a target pixel. Therefore, information on a target region and plural reference regions is obtained for the target pixel. Whether the target region matches any one of the reference regions is determined from the obtained information. Switching between weight derivation methods based on similarity between the target region and the reference region is performed according to a determined result.

Abstract: An image processing apparatus includes a setting unit, a detection unit, a matrix calculation unit, a calculation unit, an obtaining unit, a deriving unit, a patch correction unit, and a generation unit. The setting unit sets a target patch in an input image. The detection unit detects a plurality of similar patches in the input image. The matrix calculation unit calculates a covariance matrix representing correlation between pixels based on the plurality of similar patches. The calculation unit calculates eigenvalues and eigenvectors of the covariance matrix. The obtaining unit obtains a noise amount in the input image. The deriving unit derives a correction matrix based on the eigenvalues, the eigenvectors, the noise amount, and the number of similar patches. The patch correction unit corrects values of pixels in the similar patches based on the correction matrix. The generation unit generates an output image by combining the corrected similar patches.

Abstract: An image processing apparatus configured to divide an image into frequency bands and reduce noise, in which a first image includes a high-band frequency component, and a second image includes a low-band frequency component, includes: a first detecting unit configured to detect an edge in at least one of the first image and the second image; a second detecting unit configured to detect a low-contrast edge that is of a lower contrast than a contrast of the edge detected by the first detecting unit in the at least one of the first image and the second image; a compositing unit configured to composite the first image and the second image using a weighting corresponding to the edge and the low-contrast edge.

Abstract: An image processing apparatus processes image data. A first correction unit (i) corrects a pixel value of a target pixel in the image data based on a first similarity determined by comparing pixels included in a first target pixel group including the target pixel with pixels included in a first reference pixel group including a reference pixel, and (ii) produces a first corrected pixel value of the target pixel. A second correction unit (i) corrects the pixel value of the target pixel based on a second similarity determined by comparing pixels included in a second target pixel group including the target pixel with pixels included in a second reference pixel group including a reference pixel, and (ii) produces a second corrected pixel value. An output unit obtains corrected image data by using the first and second corrected pixel values based on the first similarity or the second similarity.

Abstract: An image processing apparatus includes a processing unit to perform noise reduction processing on image data and an output unit to output noise-reduced output image data. The processing unit determines a weight for a reference pixel based on a target area including a target pixel and a reference area including the reference pixel and sets, as a pixel value of the target pixel on which processed noise reduction processing has been executed, a value calculated using reference pixels and a weight corresponding to each of the reference pixels. The size of the target area equals to the size of the reference area. The output unit outputs, as the pixel value of the target pixel after noise reduction processing, either a first value, calculated when a first pixel group is set as the target area, or a second value, calculated when a second pixel group is set as the target area.

Abstract: A non-local means method is insufficient in its noise reduction effect or edge retainability due to a perfect match between blocks in a case where a reference pixel matches a target pixel. Therefore, information on a target region and plural reference regions is obtained for the target pixel. Whether the target region matches any one of the reference regions is determined from the obtained information. Switching between weight derivation methods based on similarity between the target region and the reference region is performed according to a determined result.

Abstract: A non-local means method is insufficient in its noise reduction effect or edge retainability due to a perfect match between blocks in a case where a reference pixel matches a target pixel. Therefore, information on a target region and plural reference regions is obtained for the target pixel. Whether the target region matches any one of the reference regions is determined from the obtained information. Switching between weight derivation methods based on similarity between the target region and the reference region is performed according to a determined result.

Abstract: A semiconductor device is provided that is inexpensively manufactured with variation in high-frequency characteristics suppressed. Internal matching circuit boards are disposed on at least one signal transmission path of an input-side signal transmission path between an input terminal and a semiconductor element and an output-side signal transmission path between the semiconductor element and an output terminal and is provided for matching at least between output impedance of an external circuit connected to the input terminal and input impedance of the semiconductor device or between input impedance of an external circuit connected to the output terminal and output impedance of the semiconductor device, and components are electrically connected by at least one wire causing a change exceeding an allowable value in high-frequency characteristics of the semiconductor device due to a change in wire length and are disposed in contact with each other inside a package.

Abstract: An image generation method is provided for generating an output image from an input image acquired by an image sensor that has an array of multiple pixels, each of which has arranged therein multiple sub-pixels that each receive a light beam that passes through a different pupil sub-area of an imaging optical system. The method includes a step of generating multiple parallax images that respectively correspond to the different pupil sub-areas; a step of generating multiple pixel shifted images by performing different shifting for each of the parallax images according to a virtual image forming plane of the imaging optical system; and a step of generating an output image that has a higher resolution than the resolution of the parallax images from the pixel shifted images through composition processing.

Abstract: In a case where a weighted mean value is calculated by calculating weights according to a degree of similarity based on a target pixel, the mean value is affected by noise included in the target pixel. The mean value thereby converges at a value different from a true pixel value without noise and this appears as a remaining noise. To solve this problem, an output value of the target pixel is determined based on a correlation established among a weighted mean value of multiple reference pixels, a pixel value of the target pixel, and the true value of the target pixel.

Abstract: A non-local means method is insufficient in its noise reduction effect or edge retainability due to a perfect match between blocks in a case where a reference pixel matches a target pixel. Therefore, information on a target region and plural reference regions is obtained for the target pixel. Whether the target region matches any one of the reference regions is determined from the obtained information. Switching between weight derivation methods based on similarity between the target region and the reference region is performed according to a determined result.

Abstract: Noise in RAW image data is reduced by determining parameters including pixels used as a target area and reference pixels in the RAW image data, based on color filter information for the RAW image data. The RAW image data is corrected based on the parameters thus determined.

Abstract: An image generation method is provided for generating an output image from an input image acquired by an image sensor that has an array of mnitple pixels, each of which has arranged therein multiple sub-pixels that each receive a light beam that passes through a different pupil sub-area of an imaging optical system. The method includes a step of generating multiple parallax images that respectively correspond to the different pupil sub-areas; a step of generating multiple pixel shifted images by performing different shifting for each of the parallax images according to a virtual image forming plane of the imaging optical system; and a step of generating an output image that has a higher resolution than the resolution of the parallax images from the pixel shifted images through composition processing.

Abstract: A non-local means method is insufficient in its noise reduction effect or edge retainability due to a perfect match between blocks in a case where a reference pixel matches a target pixel. Therefore, information on a target region and plural reference regions is obtained for the target pixel. Whether the target region matches any one of the reference regions is determined from the obtained information. Switching between weight derivation methods based on similarity between the target region and the reference region is performed according to a determined result.

Abstract: An image generation method is provided for generating an output image from an input image acquired by an image sensor that has an array of multiple pixels, each of which has arranged therein multiple sub-pixels that each receive a light beam that passes through a different pupil sub-area of an imaging optical system. The method includes a step of generating multiple parallax images that respectively correspond to the different pupil sub-areas; a step of generating multiple pixel shifted images by performing different shifting for each of the parallax images according to a virtual image forming plane of the imaging optical system; and a step of generating an output image that has a higher resolution than the resolution of the parallax images from the pixel shifted images through composition processing.