Abstract: This disclosure describes systems, devices, and techniques for training neural networks to identify avascular and signal reduction areas of Optical Coherence Tomography Angiography (OCTA) images and for using trained neural networks. By identifying signal reduction areas in OCTA images, the avascular areas can be detected with high accuracy, even when the OCTA images include artifacts and other types of noise. Accordingly, various implementations described herein can accurately identify avascular areas from real-world clinical OCTA images. In various implementations, a method can include identifying images of retinas. The images may include thickness images, reflectance intensity maps, and OCTA images of the retinas. Avascular maps corresponding to the OCTA images can be identified. A neural network can be trained based on the images and the avascular maps.

Abstract: State of art techniques performing image labeling of remotely sensed data are computation intensive, consume time and resources. A method and system for efficient retrieval of a target in an image in a collection of remotely sensed data is disclosed. Image scanning is performed efficiently, wherein only a small percentage of pixels from the entire image are scanned to identify the target. One or more samples are intelligently identified based on sample selection criteria and are scanned for detecting presence of the target based on cumulative evidence score Plurality of sampling approaches comprising active sampling, distributed sampling and hybrid sampling are disclosed that either detect and localize the target or perform image labeling indicating only presence of the target.

Abstract: There is provided an inspection method that includes determining, as a reference blur component, the magnitude of the blur component of a range, the range being a range where a number of first inspection images among the plurality of inspection images falling within the range is the largest among a plurality of ranges. The inspection method includes correcting, based on the reference blur component, a second inspection image among the plurality of inspection images having the magnitude of the blur component falling outside the range. The inspection method includes comparing the first inspection image and the corrected second inspection image with a reference image, the first inspection image having the magnitude of the blur component falling within the predetermined range, the reference image being generated in advance for the measurement object.

Abstract: An image processing device, including a first corrector configured to generate a first corrected image obtained by correcting a local intensity of an input image based on a complexity of an intensity distribution of the input image and an intensity value of an area around a pixel; and a second corrector configured to generate a second corrected image obtained by correcting a global intensity of the first corrected image based on a histogram of the input image.

Abstract: A system includes a sensor, a weight sensor, and a tracking subsystem. The tracking subsystem receives an image feed of top-view images generated by the sensor and weight measurements from the weight sensor. The tracking subsystem detects an event associated with an item being removed from a rack in which the weight sensor is installed. The tracking subsystem determines that a first person or a second person may be associated with the event. In response to determining that the first or second person may be associated with the event, buffer frames are stored of top-view images generated by the sensor during a time period associated with the event. The tracking subsystem then determines, using at least one of the stored buffer frames and a first action-detection algorithm, whether an action associated with the event was performed by the first person or the second person.

Abstract: Methods and apparatus to identify headlands are disclosed. A disclosed example apparatus to identify a headland of a field includes an image analyzer to generate an image from data corresponding to path information of at least one field operation performed on the field, a mask generator to generate a mask of the image based on the image, and a headland identifier to identify the headland in the field based on the mask.

Abstract: A method and apparatus for classification using a neural network. A classification apparatus includes at least one processor and a memory. The memory is configured to store a classifier and a preprocessor including a defensive noise generator. The at least one processor generates defensive noise from an input image through the defensive noise generator in the preprocessor, generates a combined image by combining the input image and the defensive noise, and outputs a classification result by inputting the combined image to the classifier.

Abstract: A system and method for detecting motion with respect an object includes providing a processor and at least one video sensor; synchronizing the at least one video sensor to a dynamic event associated with the object; recording, by the at least one video sensor, a plurality of data sets including the object, each data set of the plurality of data sets being synchronized with respect to the dynamic event associated with the object; averaging, by the processor, the plurality of data sets to provide an averaged synchronized data set; and calculating, by the processor, a motion with respect to the object based on the averaged synchronized data set.

Abstract: An information processing system, which determines a partial area used in an erectness determination which determines whether a determination target image is erect based on a determination feature in the partial area in an image having a predetermined format and a position of the partial area in a state in which the image having the predetermined format is erect, includes processor to receive input of a plurality of learning images having the predetermined format, extract area common to the plurality of learning images in an erect state as a candidate for the partial area, determine common area reliability indicative of a degree to which the candidate for the partial area is suitable as the partial area used in the erectness determination, and determine the partial area used in the erectness determination from the candidate for the partial area, based on the common area reliability.

Abstract: A system includes: a depth module including an encoder and a decoder and configured to: receive a first image from a first time from a camera; and based on the first image, generate a depth map including depths between the camera and objects in the first image; a pose module configured to: generate a first pose of the camera based on the first image; generate a second pose of the camera for a second time based on a second image; and generate a third pose of the camera for a third time based on a third image; and a motion module configured to: determine a first motion of the camera between the second and first times based on the first and second poses; and determine a second motion of the camera between the second and third times based on the second and third poses.

Abstract: The embodiments are directed to generating synthetic data. For example, in some examples, a method for creating 2D images of an object with annotation points is carried out by a processing unit. The method includes considering a synthetic 3D model of the object, named synthetic object, with annotation points correctly placed on the synthetic object. The method also includes generating several synthetic objects with different poses and different brightness conditions. Further, the method includes considering several 2D images with different backgrounds and different brightness conditions. For each 2D image, the method includes insertion of a generated synthetic object in the 2D image.

Abstract: A method of image evaluation when performing SIM microscopy on a sample includes: A) providing n raw images of the sample, which were each generated by illuminating the sample with an individually positioned SIM illumination pattern and imaging the sample in accordance with a point spread function, B) providing (S1) n illumination pattern functions, which each describe one of the individually positioned SIM illumination patterns, C) providing (S1) the point spread function and D) Carrying out an iteration method, which includes following iteration steps a) to e), as follows: a) providing an estimated image of the sample, b) generating simulated raw images, in each case by image processing of the estimated image using the point spread function and one of the n illumination pattern functions such that n simulated raw images are obtained, c) assigning each of the n simulated raw images to that of the n provided raw images which was generated by the illumination pattern that corresponds to the illumination patter

Abstract: A method is provided for inspection of an item comprising acquiring a projection image of the item using a radiation imaging system and obtaining a plurality of simulated projection images of the item or a component thereof, based on a simulation of a numerical three-dimensional model. A relative orientation of the item with respect to the imaging system is determined by comparing the projection image to the plurality of simulated images, and at least one angle of rotation is determined by taking into account a viewing angle and the relative orientation. The method further comprises moving the item and/or the imaging system in accordance with the at least one angle of rotation and acquiring a further projection image of the item.

Abstract: Disclosed herein is a method and a system for reconstructing a three-dimensional image of an object, based on stitched images of the object obtained using multiple beams.

Abstract: Apparatuses, methods, systems, and program products are disclosed for capturing an image and determining whether a switch is in a filtering position or a non-filtering position. When the switch is in the filtering position, the camera apparatus detects a switch position and determines whether the switch is in the filtering position or the non-filtering position. The camera captures the image, including audio/visual signals, while the switch is detected to be in the filtering position and modifies the image to inhibit biometric data collection from the image in response to the switch being in the filtering position when the image is captured.

Abstract: A computer system is structured to determine a density of particulate matter in a diesel particulate filter (DPF) sample. The computer system includes a processing circuit having a processor and a memory. The processing circuit is structured to generate a computed tomography (CT) scan-based image of the DPF sample; and, segment the CT scan-based image of the DPF sample into a plurality of regions. For at least one region from the plurality of regions, the processing circuit is structured to determine a density of a portion of the DPF sample corresponding to the at least one region of the CT scan-based image of the DPF sample and cause an electronic display of a user device to display the CT scan-based image including the at least one region and an indication of the density for the at least one region.

Abstract: A processor acquires a first radiographic image and a second radiographic image that include the same subject and have different S/N ratios. The processor derives a processing content of a first graininess suppression process on the first radiographic image having a higher S/N ratio of the first radiographic image and the second radiographic image and derives a processing content of a second graininess suppression process on the second radiographic image on the basis of the processing content of the first graininess suppression process. The processor performs a graininess suppression process on the second radiographic image on the basis of the processing content of the second graininess suppression process.

Abstract: Described herein are means for training a deep model to learn contrastive representations embedded within part-whole semantics via a self-supervised learning framework, in which the trained deep models are then utilized for the processing of medical imaging. For instance, an exemplary system is specifically configured for performing a random cropping operation to crop a 3D cube from each of a plurality of medical images received at the system as input; performing a resize operation of the cropped 3D cubes; performing an image reconstruction operation of the resized and cropped 3D cubes to predict the resized whole image represented by the original medical images received; and generating a reconstructed image which is analyzed for reconstruction loss against the original image representing a known ground truth image to the reconstruction loss function. Other related embodiments are disclosed.

Abstract: A method includes obtaining a source training dataset that includes a plurality of source training images and obtaining a target training dataset that includes a plurality of target training images. For each source training image, the method includes translating, using the forward generator neural network G, the source training image to a respective translated target image according to current values of forward generator parameters. For each target training image, the method includes translating, using a backward generator neural network F, the target training image to a respective translated source image according to current values of backward generator parameters. The method also includes training the forward generator neural network G jointly with the backward generator neural network F by adjusting the current values of the forward generator parameters and the backward generator parameters to optimize an objective function.

Abstract: Robust recognition systems to identify potentially identifying information that is contained within a dataset representing a system target are disclosed, together with related method. In an implementations, the systems and methods receive the dataset, identify features in the dataset that have a characteristic indicative of an expected feature, processes the identified features to yield a candidate feature, wherein the candidate feature is one feature contained within the identified features that has the highest probability of containing potentially identifying information.