Abstract: An image processing method and an image processing device are provided. The image processing method includes performing at least one of a scratch repairing step, a dead point repairing step, a denoising step and a color cast correcting step on a to-be-processed video frame. According to the embodiments of the present disclosure, it is able to repair a scratch and a dead point, remove a noise and/or correct color cast for a video frame, thereby to improve a display effect of the video frame.

Abstract: The present disclosure relates to a method comprising: providing a trained machine learning model. The trained machine learning model is configured for reconstructing images from input data. The method comprises: receiving (201) a multidimensional array comprising M dimensional acquired data; determining (205) a subset of values of at least one K selected dimension of the array; for each value of the subset determining (207) a M?K dimensional array comprising the acquired data corresponding to the value, resulting in a set of M?1 dimensional arrays; inputting (209) the set of M?K dimensional arrays to the trained machine learning model, and receive a reconstructed image from the trained machine learning model.

Abstract: The present disclosure relates to a signal processing device and an image display apparatus including the same. A signal processing device according to an embodiment of the present disclosure includes: an input interface to receive an image signal; a first image processor to generate first image frame data based on the image signal; a second image processor to generate second image frame data scaled down from the first image frame data based on the image signal; and an output interface to receive the first image frame data from the first image processor and the second image frame data from the second image processor and to output the first image frame data and the second image frame data, wherein the first image frame data output from the output interface is more delayed than the second image frame data output from the output interface. Accordingly, a timing controller may accurately and rapidly perform signal processing for a panel.

Abstract: A method, device, and system for generating medical image with localized artifacts are provided. In one aspect, the method includes receiving, by a processing unit, a medical image generated by an imaging modality, identifying presence of one or more artifacts in the medical image using one or more trained machine learning models. Herein, identifying presence of artifacts in medical image includes identifying artifacts in medical image using the trained machine learning models and determining class associated with identified artifacts in medical image using predefined classification index. The method further includes determining at least one artifact of interest from the artifacts based on a predefined criteria, localizing the at least one artifact of interest in the medical image using a post processing technique, and outputting the medical image including localized artifact of interest on a graphical user interface.

Abstract: A method for a distortion calibration of ultra-wide angle imaging apparatus is provided.

Abstract: Disclosed are a method and apparatus for integrating image enhancement and training data generation using a generative recurrent neural network model. The method of integrating image enhancement and training data generation using a generative recurrent neural network model includes: (a) receiving a target image as input, and (b) applying the target image to a trained generative recurrent neural network model to selectively generate any one of a high-quality image with enhanced image quality and a low-quality image depending on a type of the target image.

Abstract: Provided is a method for bone suppression in an X-ray image, which includes: extracting an upper contour line and a lower contour line corresponding to a bone to be suppressed from the original X-ray image; generating a first binarization image and a second binarization image based on the upper contour line and the lower contour line, respectively; generating a first distance transform image and a second distance transform image from the first binarization image and the second binarization image, respectively through distance transform; generating a compensated first X-ray image and a compensated second X-ray image by compensating a pixel value of a region which belongs to the bone by using the first distance transform image and the second distance transform image, respectively from the original X-ray image; and synthesizing the compensated first X-ray image and the compensated second X-ray image to obtain a bone-suppressed X-ray image.

Abstract: In one embodiment, a signal light state detection system includes one or more processors, an a non-transitory memory module storing computer-readable instructions. The computer-readable instructions are configured to cause the one or more processors to receive a first image of a vehicle and receiving a second image of the vehicle, wherein the second image is later in time from the first image, and generate a warped image from the first image and the second image, wherein the warped image has individual pixels of one of the first image and the second image that are shifted to locations of corresponding pixels of the other of the first image and the second image. The one or more processors further generate a difference image from the warped image and one of the first image and the second image, and determine, using a classifier module, a probability of a state of vehicle signal lights.

Abstract: The present invention provides an intelligent and a flexible video enhancement system with improved visual quality of the video for better user experience. The video enhancement system enhances various types of videos and provides a user with better visual experience based on video corresponding features and personal preferences. The video enhancement system includes a dynamic analysis and adaptive processing techniques for enhancing video quality for seamless user experience.

Abstract: The disclosure herein relates to methods and systems for localized smoke removal and color restoration of a real-time video. Conventional techniques apply the de-smoking process only on a single image, by finding the regions having the smoke, based on manual air-light estimation. In addition, regaining original colors of de-smoked image is quite challenging. The present disclosure herein solves the technical problems. In the first stage, video frames having the smoky and smoke-free video frames are identified, from the video received in the real-time. In the second stage, an air-light is estimated automatically using a combined feature map. An intermediate de-smoked video frame for each smoky video frame is generated based on the air-light using a de-smoking algorithm. In the third and the last stage, a smoke-free video reference frame is used to compensate for color distortions introduced by the de-smoking algorithm in the second stage.

Abstract: Systems and method for correcting images including artifacts due to dirty camera lenses of electronic device are disclosed.

Abstract: A computing device is configured to detect inconsistencies on a vehicle that are typically small and difficult to detect using conventional inspection techniques. Particularly, a spotlight illuminates a target area on the vehicle. The color of the spotlight is the same color as the ambient light illuminating the vehicle but is a higher intensity. Additionally, the color of the spotlight and the ambient light is selected based on the type of inconsistencies expected to be detected. Images of the target area are then captured, digitally processed, and stored in memory, such as a database, for example. Based on the processing of these images, a weighting coefficient is computed and used to generate an enhanced contrast feature map. Inconsistencies are then detected based on the enhanced contrast feature map.

Abstract: A first image collected by an image collector is acquired. A map generated based on a correspondence between a distortion rate and a field of view of the image collector is acquired. The map includes a mapping relation between a raw pixel coordinate of a pixel in a raw image collected by the image collector and a target pixel coordinate that has been corrected. A second image that has been corrected is acquired by correcting the first image based on the map.

Abstract: According to one implementation, a system for performing re-noising and neural network (NN) based image enhancement includes a computing platform having a processing hardware and a system memory storing a software code, a noise synthesizer, and an image restoration NN. The processing hardware is configured to execute the software code to receive a denoised image component and a noise component extracted from a degraded image, to generate, using the noise synthesizer and the noise component, synthesized noise corresponding to the noise component, and to interpolate, using the noise component and the synthesized noise, an output image noise. The processing hardware is further configured to execute the software code to enhance, using the image restoration NN, the denoised image component to provide an output image component, and to re-noise the output image component, using the output image noise, to produce an enhanced output image corresponding to the degraded image.

Abstract: A moving target focusing method and system based on a generative adversarial network are provided. The method includes: generating, using a Range Doppler algorithm, a two-dimensional image including at least one defocused moving target, as a training sample; generating at least one ideal Gaussian point in a position of at least one center of the at least one defocused moving target in the two-dimensional image, to generate a training label; constructing the generative adversarial network, the generative adversarial network includes a generative network and a discrimination network; inputting the training sample and the training label into the generative adversarial network to perform repeated training until an output of the generative network reaches a preset condition, to thereby obtain a trained network model; and inputting a testing sample into the trained network model, to output a moving target focused image.

Abstract: An image segmentation system and display is disclosed. The system may be operated or configured to generate a segmentation of a member from an image. The image and/or the segmentation may be displayed for viewing by a user.

Abstract: Disclosed are techniques for image processing that mitigate artifacts that otherwise appear when an IQVR camera moves with respect to a scene between image frames.

Abstract: Methods, apparatus, systems and articles of manufacture to reconstruct scenes using convolutional neural networks are disclosed. An example apparatus includes a sensor data acquirer to acquire ground truth data representing an environment, an environment detector to identify an environmental characteristic of the environment, a synthetic database builder to apply noise to the ground truth data to form a training set, a model builder to train a machine learning model using the training set and the ground truth data, and a model adjustor to modify the machine learning model to include residual OR-gate connections intermediate respective layers of the machine learning model. The synthetic database builder is further to store the machine learning model in association with the environmental characteristic of the environment.

Abstract: A single image deraining method is proposed. A wavelet transforming step processes an initial rain image to generate an i-th stage low-frequency rain image and a plurality of i-th stage high-frequency rain images. An image deraining step inputs the i-th stage low-frequency rain image to a low-frequency deraining model to output an i-th stage low-frequency derain image. A first inverse wavelet transforming step recombines the n-th stage low-frequency derain image with the n-th stage high-frequency derain images to form an n-th stage derain image. A weighted blending step blends a (n?1)-th stage low-frequency derain image with the n-th stage derain image to generate a (n?1)-th stage blended derain image. A second inverse wavelet transforming step recombines the (n?1)-th stage high-frequency derain images with the (n?1)-th stage blended derain image to form a (n?1)-th stage derain image, and sets n to n?1 and repeats the last two steps.

Abstract: Changes in ambient light intensity are detected by processing image data from an image capture device to determine, iteratively, a signal-to-noise ratio of the image data. The signal-to-noise ratio is a ratio of average to variance of pixel values for some of the pixels forming the image. A control output is generated, based on the signal-to-noise ratio, that is responsive to changes in ambient light intensity. The control output is used control a light source of a vehicle.

Abstract: A computing device stores a predictive model generated by a neural network training engine. The computing device receives first and second two-dimensional (2D) thermal images comprising temperature measurements from respective first and second infrared (IR) sensors. The first and second images have the same size. An image capturing visual field of the second IR sensor partially overlaps with an image capturing visual field of the first IR sensor. The computing device executes a neural network using a predictive model for generating outputs based on inputs. The inputs comprise the temperature measurements of the first and second images. The outputs comprise horizontal and vertical shifts defining a translation of the second image with respect to the first image. An overlapping area in the first image, having a rectangular shape and overlapping with the second image, is determined using the horizontal and vertical shifts.

Abstract: A method of detecting a defect in an electronic display includes taking a digital image of the electronic display; and electronically determining if an area of the electronic display is defective based on data of the digital image.

Abstract: A solenoid valve has a valve portion and a solenoid portion. The valve portion includes a sleeve and a spool. The solenoid portion includes a yoke, a plunger, a stator core, and a base portion. The base portion includes a first inner diameter portion, a second inner diameter portion, and a connection surface that connects the first inner diameter portion and the second inner diameter portion to each other in a radial direction. The first inner diameter portion is fastened to a first outer peripheral surface of the sleeve. The second inner diameter portion is arranged radially outside a second outer peripheral surface of a magnetic attraction core. The base portion is in contact with the yoke. The connection surface and a sleeve end surface of the sleeve adjacent to the solenoid portion in the axial direction are spaced from each other in the axial direction.

Abstract: Various examples are provided for distortion rectification and fingerprint crossmatching. In one example, a method includes selecting an electronic, perspective distorted fingerprint sample; and generating an unwarped fingerprint sample by rectifying perspective distortions from the perspective distorted fingerprint sample by application of an unwarping transformation. Parameters of the unwarping transformation can be determined by a deep convolutional neural network (DCNN) trained on a database comprising contactless fingerprint samples suffering from perspective distortions.

Abstract: Provided are an image processing apparatus and an operation method of the image processing apparatus. The image processing apparatus includes a memory storing one or more instructions, and a processor configured to execute the one or more instructions stored in the memory to, by using one or more convolution neural networks, extract target features by performing a convolution operation between features of target regions having same locations in a plurality of input images and a first kernel set, extract peripheral features by performing a convolution operation of features of peripheral regions located around the target regions in the plurality of input images and a second kernel set, and determine a feature of a region corresponding to the target regions in an output image, based on the target features and the peripheral features.

Abstract: A method for edge identification of a self-shadowing object based on a visual camera and a vehicle includes: collecting image data of an object to be identified, and performing differentiation processing; according to a preset threshold, performing three-value processing to the image that has been differentiation-processed, to acquire a three-value-processed image including positive-direction boundary pixels and negative-direction boundary pixels; according to the positive-direction boundary pixels and the negative-direction boundary pixels, acquiring a positive-direction straight linear segment and a negative-direction straight linear segment that represent boundary trends of the object to be identified; if the straight linear segments do not create a target object that satisfy a predetermined condition, determining the first boundary according to the predetermined condition; according to the first boundary, determining the second boundary; and by using positions of the first boundary and the second boundary

Abstract: A display system for a welding mask that includes a darkening filter layer, a light field camera system to capture a light field, and an optical image stabilization subsystem. Captured light fields are displayed via a light field display within the welding helmet without risk of overexposure of ultraviolet radiation to the operator. In some examples, dual light field displays are used to display different images to each eye of the operator.

Abstract: A three-dimensional holographic display device includes a light emitting diode (LED) array including a plurality of light sources controlled to sequentially output light according to a preset pattern, a lens configured to refract light incident from the LED array, a spatial light modulator (SLM) configured to modulate light incident from the lens, and a processor configured to generate a plurality of holographic signals each comprising depth information adjusted according to an arrangement location of each of the plurality of light sources, and for each of the plurality of light sources, control the SLM to modulate the light based on a holographic signal corresponding to the light source.

Abstract: The present disclosure relates to systems and methods for image reconstruction. The system may obtain an initial image to be processed. The system may generate a reconstructed image by performing a plurality of iteration steps on the initial image. Each of the plurality of iteration steps may include a first optimization operation and at least one second optimization operation. The first optimization operation and the at least one second optimization operation may be executed sequentially. The first optimization operation may include receiving an image to be processed in the iteration step and determining an updated image by preliminarily optimizing the image to be processed. The at least one second optimization operation may include determining an optimized image by reducing interference information of the updated image and designating the optimized image as a next image to be processed in a next iteration step. The interference information may include noise information and/or artifact information.

Abstract: A method for concealing data in an image or a video stream inside a compression chain, being implemented by a computer and including a structuring and processing phase, during which at least one image is structured into blocks including coefficients, a phase of converting and quantifying blocks so as to generate converted and quantified coefficients, with entropic coding intended to code the converted and quantified coefficients. The method includes a step of concealing data, during which bits of the data are concealed by modifying converted and quantified coefficients located in a high frequency zone of at least some of the blocks that relate to the luminance component of the video stream and that are intra-coded. This modification is performed after the quantification but before the entropic coding.

Abstract: An apparatus includes a control unit configured to control a position of a focus lens of the apparatus, a setting unit configured to change a threshold for detecting a luminance change in each pixel in a predetermined range, an acquisition unit configured to acquire a signal indicating a pixel in which a luminance change that is more than or equal to the threshold has occurred, and a determination unit configured to determine an in-focus position of the focus lens based on the signal detected depending on the position of the focus lens and the set threshold.

Abstract: The present disclosure provides a method and device for correcting a scanned image, and an image scanning system, and relates to the field of image scanning. The method includes obtaining a scanned image of a scanned object, detecting one or more reference objects from the scanned image, determining a deformation parameter of each reference object of the one or more reference objects based on preset a standard parameter of the each reference object, and correcting the scanned image based on the deformation parameters of the one or more reference objects.

Abstract: An AR calibration system for correcting AR headset distortions. A calibration image is provided to an external screen and viewable through a headset reflector, and an inverse of the calibration image is provided to a headset display, reflected off the reflector and observed by a camera of the system while it is simultaneously observing the calibration image on the external screen. One or more cameras are located to represent a user's point of view and aligned to observe the inverse calibration image projected onto the reflector. A distortion mapping transform is created using an algorithm to search through projection positions of the inverse calibration image until the inverse image observed by the camera(s) cancels out an acceptable portion of the calibration image provided to the external screen as observed through the reflector by the camera, and the transform is used by the headset, to compensate for distortions.

Abstract: Certain embodiments involve techniques for efficiently estimating denoising kernels for generating denoised images. For instance, a neural network receives a noisy reference image to denoise. The neural network uses a kernel dictionary of base kernels and generates a coefficient vector for each pixel in the reference image such that the coefficient vector includes a coefficient value for each base kernel in the kernel dictionary, where the base kernels are combined to generate a denoising kernel and each coefficient value indicates a contribution of a given base kernel to a denoising kernel. The neural network calculates the denoising kernel for a given pixel by applying the coefficient vector for that pixel to the kernel dictionary. The neural network applies each denoising kernel to the respective pixel to generate a denoised output image.

Abstract: A method of video decoding performed in a video decoder includes receiving compressed geometric data corresponding to a three dimensional (3D) space. The method further includes generating a resolution occupancy map. The method further includes reconstructing a point cloud using the compressed geometric data and the resolution occupancy map. The method further includes performing a smoothing operation on one of the (i) reconstructed point cloud and (ii) the generated resolution occupancy map.

Abstract: In various embodiments, a prediction application computes a quality score for re-constructed visual content that is derived from visual content. The prediction application generates a frame difference matrix based on two frames included in the re-constructed video content. The prediction application then generates a first entropy matrix based on the frame difference matrix and a first scale. Subsequently, the prediction application computes a first value for a first temporal feature based on the first entropy matrix and a second entropy matrix associated with both the visual content and the first scale. The prediction application computes a quality score for the re-constructed video content based on the first value, a second value for a second temporal feature associated with a second scale, and a machine learning model that is trained using subjective quality scores. The quality score indicates a level of visual quality associated with streamed video content.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for receiving mission specific data. Converting the mission specific data into attribute classes recognizable by image recognition classifiers, where the image recognition classifiers are pre-trained to detect objects corresponding to the attributed classes within digital images. Obtaining, from a wearable camera system, low resolution images of a scene and high resolution images of the scene. Detecting, within the low resolution images using a low resolution image classifier, an object that corresponds to one of the attribute classes. In response to detecting the object that corresponds to one of the attribute classes, providing, for presentation to a user of the wearable camera system by a presentation system, an first alert indicating a potential detection of the suspect, and providing the high resolution images to a high resolution image classifier.

Abstract: A method for determining whether or not a transparent protective cover of a video camera comprising a lens-based optical imaging system is partly covered by a foreign object is disclosed. The method comprises: obtaining (402) a first captured image frame captured by the video camera with a first depth of field; obtaining (404) a second captured image frame captured by the video camera with a second depth of field which differs from the first depth of field; and determining (406) whether or not the protective cover is partly covered by the foreign object by analysing whether or not the first and second captured image frames are affected by presence of the foreign object on the protective cover such that the difference between the first depth of field and the second depth of field results in a difference in a luminance pattern of corresponding pixels of a first image frame and a second image frame.

Abstract: The present disclosure discloses an image haze removal method and apparatus, and a device. The method includes: acquiring a hazy image to be processed; and obtaining a haze-free image corresponding to the hazy image by inputting the hazy image into a pre-trained haze removal model. The present disclosure uses the residual dual attention fusion modules as basic modules of the neural network, so that each feature map can obtain pixel features while enhancing the global dependence, thus improving the image dehazing effect.

Abstract: Systems and methods are disclosed for image signal processing. For example, methods may include receiving a current image of a sequence of images from an image sensor; combining the current image with a recirculated image to obtain a noise reduced image, where the recirculated image is based on one or more previous images of the sequence of images from the image sensor; determining a noise map for the noise reduced image, where the noise map is determined based on estimates of noise levels for pixels in the current image, a noise map for the recirculated image, and a set of mixing weights; recirculating the noise map with the noise reduced image to combine the noise reduced image with a next image of the sequence of images from the image sensor; and storing, displaying, or transmitting an output image that is based on the noise reduced image.

Abstract: A system, and method of operating the same detects moving targets in images and performs image turbulence correction. The system includes an automatic target recognizer (ATR) system including a database. The ATR includes a feature extractor and processor arranged to detect a plurality of reference features associated with targets within image frames, and calculate a position of the plurality of reference features. The system includes an image processor arranged to receive the position, demosaic the image frames into a plurality of video tiles, iteratively process the video tiles for turbulence correction to generate turbulence corrected video tiles associated with acquired targets; convert the turbulence corrected video tiles into a single video frame tile including turbulence degradation correction; and mosaic each of the single video frame tiles to generate a full field of view turbulence corrected video stream.

Abstract: A non-uniformity correction (NUC) calibration method comprises obtaining image data for a plurality of images with an image sensor, wherein each image in the plurality of images is obtained at a different respective global pixel gain setting and global expose in the image sensor; and using the image data for non-uniformity correction calibration to compute pixel NUC values for the pixels in the image sensor. The method can further include storing the pixel NUC values and obtaining further image data corrected by the stored pixel NUC values. In embodiments, the method can include moving a platform based on the further image data. In certain embodiments, the platform can be a guided munition.

Abstract: A method performed with an image sensor having a pixel array. At least one frame of a scene may be obtained using the pixel array. At least one region of interest (ROI) is identified within the frame. Subsequent frames of the scene are obtained, which involves controlling the pixel array to perform high resolution imaging with respect to the at least one ROI and low resolution imaging using analog binning with respect to remaining regions of the frames.

Abstract: A method may include obtaining sensor data from one or more LiDAR units and determining a point-cloud corresponding to the sensor data obtained from each respective LiDAR unit. The method may include aggregating the point-clouds as an aggregated point-cloud and generating an initial proposal for a two-dimensional ground model made of multiple grid blocks. The method may include filtering out unrelated raw data points from each grid block of the plurality of grid blocks to generate a filtered point-cloud matrix. The method may include identifying one or more surface-points and one or more object-points included in the filtered point-cloud matrix and generating an array of extracted objects based on the object-points.

Abstract: Image signal processing includes obtaining two or more image signals from a first hyper-hemispherical image sensor, where each of the two or more image signals has a different exposure and obtaining two or more image signals from a second hyper-hemispherical image sensor, where each of the two or more image signals has a different exposure. Image signal processing includes generating an exposure compensated image based on a gain value applied to an exposure level of a first image and a gain value applied to an exposure level of a second image. Image signal processing further includes performing high dynamic range (HDR) processing on the exposure compensated image. The HDR processing may be performed on a high a frequency portion of the exposure compensated image.

Abstract: A control apparatus includes a receiving unit that receives a captured image obtained by capturing an image in an optical finder of an image capture device in a state in which AF frames are displayed on the optical finder, a display unit that displays the captured image and a UI (user interface) for selecting at least one AF frame from AF frames displayed on the captured image, and a transmitting unit that transmits, to the image capture device, information for specifying the at least one AF frame selected via the UI.

Abstract: Examples of the present disclosure describe systems and methods for detecting and remediating compression artifacts in multimedia items. In example aspects, a machine learning model is trained on a dataset related to compression artifacts and non-compression artifacts. Input data may then be collected by a data collection module and provided to a pattern recognition module. The pattern recognition module may extract visual and audio features of the multimedia item and provide the extracted features to the trained machine learning model. The trained machine learning model may compare the extracted features to the model, and a confidence value may be generated. The confidence value may be compared to a confidence value threshold. If the confidence value is equal to or exceeds the confidence threshold, then the input data may be classified as containing at least one compression artifact. Remedial action may subsequently be applied (e.g., boosting the system with technical resources).

Abstract: Devices, systems and methods for video processing are described. In an exemplary aspect, a method for video processing includes performing a conversion between a coded representation of a video comprising one or more video regions and the video. The coded representation includes first side information that provides a clipping parameter for filtering a reconstruction of a video unit of a video region using a non-linear adaptive loop filter, and wherein the first side information is signaled together with second side information indicative of filter coefficients used in the non-linear adaptive loop filter.

Abstract: An image processing method includes: receiving specific data of an image component of input image for a specific color encoding channel wherein the specific data of the image component is formed by the image component's multiple raw values having identical image characteristics; spatially classifying pixel units within the input image into a plurality of blocks according to a specific block size; for each block, obtaining values of the image component of pixel units within each block to calculate the average value of the image component of the each block; generating interpolated values of the image component of an interpolated image according to values of the image component corresponding to the plurality of blocks; and, blending the interpolated values of the image component of the interpolated image with original/raw values of the image component of the input image to generate an output image having modified values of the image component.

Abstract: A method, system, and computer program product for determining selection parameters for filtering algorithms using environmental data of images. The method may include receiving an image. The method may also include analyzing the image using at least image processing. The method may also include identifying, based on the analyzing, image data and environmental data of the image. The method may also include inputting the image data and the environmental data into a machine learning algorithm, where the machine learning algorithm includes mapped relationships between at least the environmental data and selection parameters. The method may also include predicting, using the machine learning algorithm, optimal selection parameters for the image. The method may also include applying the optimal selection parameters to a filtering algorithm for the image. The system and computer program product may include similar steps.