Abstract: The invention relates to a method for processing images obtained by a diffraction detector, of a crystalline or polycrystalline material, in which a first image of the material is acquired in a state of reference as well as a second image of the material in a deformed state. The invention is characterised in that, in a calculator, during a first step (E6, E12), a current elastic deformation gradient tensor Fe is given a value determined by calculation, during a second step (E7), the current displacement field induced by the tensor Fe is calculated, during a third step (E8), third digital values of a deformed image {hacek over (g)}(x)=g(x+u(x)) corrected by the current displacement field are calculated, and during an iterative algorithm, iterations of the second and third steps (E12, E7, E8) are carried out on modified values of the tensor r Fe until a convergence criterion is met in relation to the correction to the current value of Fe.

Abstract: Provided are an image processing method and device, a mobile terminal, and a storage medium. The method includes: acquiring a first image output by an image sensor; responsive to that a first signal value of the first image after being processed exceeds a saturation value of a first bit width at a preset image processing stage, recording the first signal value of the processed first image as a second signal value with a second bit width, the second bit width being greater than the first bit width; and mapping the second signal value recorded with the second bit width to a third signal value recorded with the first bit width to obtain a second image. Through the method, an operation of increasing a bit width can be used to retain more information of an image and improve the quality of the image, and an implementation is simple and effective.

Abstract: A machine learning model can be trained using a first set of degraded images for each of a plurality of combinations and corresponding reference images, where a number of degraded images in the first set corresponding to a particular combination of the plurality of combinations is selected in accordance with a probability value associated with the particular combination. A validation process can be used to determine a loss value for each of the plurality of combinations of degradations. Updates to the probability values associated with the plurality of combinations can be calculated based on the loss values. The machine learning model can be updated using a second set of degraded images for each of the plurality of combinations, and the corresponding reference images, where a number of degraded images in the second set corresponding to the particular combination is selected based on the updated probability value.

Abstract: In one embodiment, a computing system may receive an uncompressed image from a camera. The computing system may generate a compressed image by performing a compression process on the uncompressed image, wherein a decompressed image may be generated as a byproduct during the compression process. The computing system may send the decompressed image to a machine-learning model that was trained using decompressed images. The computing system may generate, by the machine-learning model, an output based on the decompressed image. The computing system may provide operational instructions to a vehicle based on the output.

Abstract: An ophthalmological image processing apparatus acquires a plurality of images of a subject eye photographed in a scanning-type imaging optical system, sets any one of the plurality of images as a template, sets corresponding points or corresponding regions between an image of the subject eye and the template at a plurality of positions of each of the image of the subject eye and the template, calculates a movement amount of each of the corresponding points or each of the corresponding regions, and corrects a distortion of the image of the subject eye with respect to the template based on the movement amount of each of the corresponding points or each of the corresponding regions.

Abstract: An image processing method for generating an estimated image in which a defocus blur shape in a captured image is corrected includes a first step of acquiring input data including the captured image and shape designating information that designates a defocus blur shape in the estimated image, and a second step of inputting the input data to a machine learning model and of generating the estimated image.

Abstract: Embodiments are disclosed for generating lens blur effects. The disclosed systems and methods comprise receiving a request to apply a lens blur effect to an image, the request identifying an input image and a first disparity map, generating a plurality of disparity maps and a plurality of distance maps based on the first disparity map, splatting influences of pixels of the input image using a plurality of reshaped kernel gradients, gathering aggregations of the splatted influences, and determining a lens blur for a first pixel of the input image in an output image based on the gathered aggregations of the splatted influences.

Abstract: A non-transitory computer-readable recording medium stores therein a learning program that causes a computer to execute a process including: generating a shadow image including a shadow according to a state of ultrasound reflection in an ultrasound image; generating a combined image by combining the ultrasound image and the shadow image; inputting, into a first decoder and a second decoder, an output acquired from an encoder in response to inputting the combined image into the encoder; and executing training of the encoder, the first decoder, and the second decoder, based on: reconfigured error between an output image of a coupling function and the combined image, the coupling function being configured to combine a first image output from the first decoder with a second image output from the second decoder, and an error function between an area in the first image and the shadow in the shadow image.

Abstract: Methods and systems for aligning images of a specimen are provided. One method includes reducing noise in a test image generated for a specimen by an imaging subsystem thereby generating a denoised test image. The method also includes detecting one or more patterned features in the denoised test image extending in at least a horizontal or vertical direction. In addition, the method includes designating an area of the denoised test image in which the detected one or more patterned features are located as a region of interest in the denoised test image. The method further includes aligning the denoised test image to a reference image for the specimen using only the region of interest in the denoised test image and a corresponding area in the reference image.

Abstract: A method for determining and suppressing Moiré pattern, and a circuit system thereof are provided. In the method, a brightness value of pixels of an image can be obtained. For each of the pixels of the image, a detection window is provided for calculating a Moiré pattern response value of a plurality of critical pixels and corresponding adjacent pixels. The critical pixels within the detection window are selected for determining a type of Moiré pattern. After, it is to compare the brightness values of the critical pixels and the corresponding adjacent pixels within the detection window. The comparison results can be used to determine the brightness characteristics of the pixels through a statistical method. Moiré pattern response value and the statistics are used to determine type and position of the Moiré pattern. A color noise suppression process is performed on the pixels to be determined as Moiré pattern.

Abstract: Systems and methods are provided for evaluating an eye using retinal fluid volumes to provide a clinical parameter. An optical coherence tomography (OCT) image of an eye of a patient is obtained. The OCT image is segmented to produce a total retinal volume and one or both of a subretinal fluid volume and an intraretinal fluid volume for a region of interest within the eye. A metric is generated as a function of the total retinal volume and one or both of the subretinal fluid volume and the intraretinal fluid volume. A clinical parameter for the patient is determined from the metric. The determined clinical parameter is provided to a user at a display.

Abstract: Teeth are screened for periodontal disease using digitized images manipulated and annotated on a processor. A digitized radiographic image of a tooth shows locations of a bone boundary and a cemento-enamel junction (CEJ) of the tooth. The digitized radiographic image is marked on the processor with a location on the bone boundary and with a pair of CEJ points at opposite ends of the CEJ visible in the radiograph. A ratio between (a) a distance between the bone boundary location and the adjacent CEJ point as numerator and (b) a distance between the CEJ points as denominator is calculated on the processor and compared with a threshold ratio-value for a corresponding tooth from a database accessible by the processor. A calculated ratio-value which is greater than the database threshold ratio-value is indicative of periodontal disease in the tooth. The probability of the correct diagnostic decision is determined by the relative magnitude of the calculated ratio-value and the threshold ratio-value.

Abstract: A method for providing an optimum subtraction data set includes: receiving first image data sets acquired by a medical imaging device and which map an object under examination within a first time phase; receiving at least one second image data set acquired by the same or another medical imaging device and which maps a change in the object under examination within a second time phase; dividing the at least one second image data set into a plurality of image regions; generating subtraction image regions for the plurality of image regions; determining an image quality parameter for each subtraction image region; determining an optimum subtraction image region for each image region of the plurality of image regions of the at least one second image data set by comparing the image quality parameters; generating the optimum subtraction data set from the optimum subtraction image regions; and providing the optimum subtraction data set.

Abstract: State of art systems in the domain of dynamism detection fail to estimate noise if depth images are being collected as input for the dynamism detection, as the noise in depth images depend on the scene being captured. Disclosed herein are method and system for determining dynamism by processing depth image of a scene. The system models depth sensor noise as ergodic stochastic process by determining that distribution estimated at each reference pixel from a plurality of neighborhood pixels in a reference image being processed is statistically same as a distribution estimated from evolution of the reference pixel over the time. After modeling the depth sensor noise in this manner, the same is eliminated/removed from the reference image, which is then processed to estimate divergence at each pixel based on temporal and spatial distribution built at pixel level in the reference image, and in turn determines dynamism in the scene.

Abstract: A face recognition monitoring system based on spectrum and multi-band fusion, including a spectrum camera, a first module for acquiring a face spectral image, a second module for preprocessing data of the face spectral image, a face spectral image database and a third module for recognizing the face spectral image. The spectrum camera includes an optical lens and a silicon-based detector. The silicon-based detector includes a photoelectric conversion substrate and a filter film arranged thereon. The filter film includes N units each including a visible spectrum sensing area, a near-infrared spectral image sensing area and a RGGB image acquisition area. The N units cover all pixels on the photoelectric conversion substrate. A recognition method using the above system is also provided.

Abstract: A radar image processing device includes: an estimation unit 1 for setting, as a target pixel, each pixel of a two-dimensional map image, and estimating to which type each target pixel belongs among a pixel caused by the main lobe, a pixel due to the sidelobe, and any other pixel; a pixel value replacement unit 2 for replacing the pixel value of the pixel caused by the main lobe and the pixel value of the pixel due to the sidelobe with pixel values generated in pixel value interpolation processing based on the type of each pixel estimated by the estimation unit 1 to generate a first corrected image; a speckle noise suppression unit 3 for applying speckle noise suppression processing to the first corrected image to generate a second corrected image; and an output image generation unit 4 for generating an output image, in which speckle noise and sidelobes are suppressed, by using the two-dimensional map image, the second corrected image, and the type of each pixel.

Abstract: In an image processing system according to the present disclosure, image data after conversion is obtained by converting compressed image data by using a neural network, in which the neural network is generated by performing learning based on first image data and second image data obtained by lossy-compressing the first image data.

Abstract: An image decoding method according to the present specification comprises the steps of: deriving transform coefficients through inverse quantization on the basis of quantized transform coefficient for a target block; deriving modified transform coefficients on the basis of inverse reduced secondary transform (RST) of the transform coefficients; and generating a reconstructed picture on the basis of residual samples for the target block on the basis of an inverse primary transform of the modified transform coefficients, wherein the step of deriving the modified transform coefficients is characterized in deriving 16 modified transform coefficients by applying a transform kernel matrix to 8 transform coefficients in a 4×4 region of the target block.

Abstract: Techniques and systems are described for automatic artifact removal in a digital image. A segmentation map is generated that describes a magnitude of difference among pixels in a digital image. Contours may be generated that describe boundaries of objects described in the segmentation map. The contours may be filtered according to two-dimensional and three-dimensional cues to identify contours corresponding to artifacts in the digital image. For each contour corresponding to an artifact, an object mask and a sampling mask may be generated. The object mask and the sampling mask may be utilized as part of a content filling operation upon the digital image to remove the artifact, and a corrected digital image is generated that does not include the artifact.

Abstract: A method and a device for encoding/decoding images are disclosed. The method for encoding images comprises the steps of: deriving a scan type of a residual signal for a current block according to whether or not the current block is a transform skip block; and applying the scan type to the residual signal for the current block, wherein the transform skip block is a block to which transform for the current block is not applied and is specified on the basis of information indicating whether or not transform for the current block is to be applied.