Abstract: Provided are an image processing apparatus, a medical imaging apparatus, and an image processing program that remove noise from an image having different noise levels depending on regions in the image at a low calculation cost, and enable high quality according to a preference of a reader. A plurality of image generators receive measurement data or a captured image obtained by an image data acquisition apparatus and generate different images for a same imaging range. An image selection and combination unit selects different image regions from a plurality of images generated by the plurality of image generators according to a predetermined region selection pattern, and generates one image by combining the images of the selected image regions.

Abstract: Embodiments of the present application provide a video control method, a video conference terminal and a multi-point control unit (MCU). The method comprises: performing human facial recognition on a video code stream and determining a motion quantity of the video code stream; sorting identifiers P and motion quantities R respectively corresponding to a plurality of video code streams; and selecting video code streams corresponding to video conference terminals ranked top M for synthesis and output and/or selecting a video conference terminal corresponding to a video code stream ranked first as a broadcast source.

Abstract: Approaches for analyzing an input image and providing one or more outputs related to the input image are provided. In accordance with an exemplary embodiment, an input image may be received and analyzed, using a trained machine learning model, to generate an inference related to the image. Based, at least in part, upon the generated inference, one or more reports related to the inference can be generated and provided for presentation on a user device. A user can interact with the report in a conversational manner with the computer system to generate additional reports or insights related to the input image.

Abstract: A system and method configured to better identify patient-specific anatomical landmarks, measure anatomical parameters and features, and predict the patient's need for surgery within a predetermined time period. In some embodiments, the system and method is configured to predict the likelihood or risk that a patient will require total hip arthroplasty. In some embodiments, the present invention includes machine learning technology Some embodiments of the present invention include a first ML machine configured to received medical images as inputs and identify anatomical landmarks as outputs; a measurement module to measure joint space width, hip dysplasia angles, and/or leg length differential; and a second ML machine configured to receive the anatomical measurements and patient demographic data as inputs and produce a risk or likelihood that the patient will require surgery within a certain time frame.

Abstract: A system and a method for image-based crop identification are disclosed. The image-based crop identification system includes a database, a communication module and a model library. The database stores sample aerial data and annotated aerial data. The communication module is coupled to the database, and is configured to provide the sample aerial data to a user and receive the annotated aerial data from the user. The model library is coupled to the database, and is configured to obtain the annotated aerial data, train a crop classification model based on the annotated aerial data, and provide the trained crop classification model for subsequent crop identification. The annotated aerial data include determination of the type of the crop appearing in the sample aerial data.

Abstract: The disclosure is directed to, among other things, systems and methods for troubleshooting equipment installations using machine learning. Particularly, the systems and methods described herein may be used to validate an installation of one or more devices (which may be referred to as “customer premises equipment (CPE)” herein as well) at a given location, such as a customer's home or a commercial establishment. As one non-limiting example, the one or more devices may be associated with a fiber optical network, and may include a modem and/or an optical network terminal (ONT). However, the one or more devices may include any other types of devices associated with any other types of networks as well.

Abstract: Provided is a control method for a system for determining a lesion obtained via real-time image. The control method comprises: an endoscope device obtaining a stomach endoscopy image; the endoscope device transmitting the obtained stomach endoscopy image to a server; the server determining a lesion included in the stomach endoscopy image, by inputting the stomach endoscopy image into a first artificial intelligence model; when it is determined that a lesion is detected in the stomach endoscopy image, the server obtaining an image including the lesion and transmitting the image to a database of the server; the server determining the type of the lesion included in the image, by inputting the image into a second artificial intelligence model; and when it is determined that a lesion is detected in the stomach endoscopy image, a display device displaying a UI for guiding the location of the lesion in the stomach endoscopy image.

Abstract: A system is disclosed that includes an optical tracking device and a surgical computing device. The optical tracking device includes a structured light module and an optical module that includes an image sensor and is spaced from the structured light module at a known distance. The surgical computing device includes a display device, a non-transitory computer readable medium including instructions, and processor(s) configured to execute the instructions to generate a depth map from a first image captured by the image sensor during projection of a pattern into a surgical environment by the structured light module. The pattern is projected in a near-infrared (NIR) spectrum. The processor(s) are further configured to execute the stored instructions to reconstruct a 3D surface of anatomical structure(s) based on the generated depth map. Additionally, the processor(s) are configured to execute the stored instructions to output the reconstructed 3D surface to the display device.

Abstract: Systems and methods are disclosed for generating synthetic medical images, including images presenting rare conditions or morphologies for which sufficient data may be unavailable. In one aspect, style transfer methods may be used. For example, a target medical image, a segmentation mask identifying style(s) to be transferred to area(s) of the target, and source medical image(s) including the style(s) may be received. Using the mask, the target may be divided into tile(s) corresponding to the area(s) and input to a trained machine learning system. For each tile, gradients associated with a content and style of the tile may be output by the system. Pixel(s) of at least one tile of the target may be altered based on the gradients to maintain content of the target while transferring the style(s) of the source(s) to the target. The synthetic medical image may be generated from the target based on the altering.

Abstract: A method for judging freshness of cultured fish product based on eye image recognition is provided, which relates to the field of image processing and includes the following steps: obtaining eye area and eye center point of each cultured fish product; obtaining a first data category and a second data category of each gray scale change sequence according to each gray scale change sequence of each eye area; calculating a first mean value and a second mean value of each gray scale change sequence; obtaining the fish eye turbidity of each eye area according to the first mean value and the second mean value; obtaining the fish eye plumpness of each eye area according to the first data category and the second data category; and obtaining the freshness of each cultured fish product according to the fish eye turbidity and fish eye plumpness in each eye area.

Abstract: A background image generator (21) stores, in a storage device (3), a skeleton image obtained by calculating a feature quantity for each of multiple first thermal images (Din1) obtained by imaging by a thermal image sensor (1) in the same field of view or multiple sorted images (Dc) generated from the first thermal images, generating an average image from the first thermal images or sorted images, sharpening the average image, and then extracting a skeleton component. An image corrector (22) corrects, by using the skeleton image stored in the storage device (3), a second thermal image (Din2) obtained by imaging by the thermal image sensor in the same field of view as the first thermal images, thereby generating a corrected thermal image. It is possible to generate a sharp thermal image with a high S/N ratio.

Abstract: Techniques for training a first electronic neural network classifier to identify a presence of a particular property in a novel supra-image while ignoring a spurious correlation of the presence of the particular property with a presence of an extraneous property are presented.

Abstract: Systems and methods for providing a novel framework to simulate the appearance of pathology on patients who otherwise lack that pathology. The systems and methods include a “simulator” that is a generative adversarial network (GAN). Rather than generating images from scratch, the systems and methods discussed herein simulate the addition of diseases-like appearance on existing scans of healthy patients. Focusing on simulating added abnormalities, as opposed to simulating an entire image, significantly reduces the difficulty of training GANs and produces results that more closely resemble actual, unmodified images. In at least some implementations, multiple GANs are used to simulate pathological tissues on scans of healthy patients to artificially increase the amount of available scans with abnormalities to address the issue of data imbalance with rare pathologies.

Abstract: An image processing method is provided for displaying cells from a plurality of pathology images or overall images. A respective overall image represents a respective patient tissue sample or a respective patient cell sample.

Abstract: The technology disclosed introduces two types of neural networks: “master” or “generalists” networks and “expert” or “specialists” networks. Both, master networks and expert networks, are fully connected neural networks that take a feature vector of an input hand image and produce a prediction of the hand pose. Master networks and expert networks differ from each other based on the data on which they are trained. In particular, master networks are trained on the entire data set. In contrast, expert networks are trained only on a subset of the entire dataset. In regards to the hand poses, master networks are trained on the input image data representing all available hand poses comprising the training data (including both real and simulated hand images).

Abstract: A method for capturing digital data for fabricating a dental splint involves displaying a GUI on a display of a smartphone that provides an alignment feature for a user to align a camera of the smartphone to a first position that captures teeth of a person, receiving digital video of the teeth, overlaying the alignment feature on the digital video of the teeth on the display, moving the alignment feature on the screen in a manner that causes the user to move the smartphone relative to the teeth to maintain alignment with the alignment feature, capturing digital image information of the teeth as the alignment feature is moved, the captured digital image information including depth information, and transmitting the captured digital image information, including the depth information, from the smartphone for use in fabricating a dental splint.

Abstract: A hydroponic system is managed by obtaining hydroponic data using a plurality of sensors in a hydroponic cultivator. Data from the plurality of sensors is communicated to a data aggregator, which renders the sensed data from the plurality of sensors in a predetermined data format, as aggregator output data. The aggregator output data is communicated to a fog computing unit, which, in turn, communicates with outside computing routed over the Internet backbone or “cloud”. The fog computing unit is used to control operation of the hydroponic cultivator based on direct control commands executed through the fog control unit and data inputs obtained by communications with outside computing routed over the Internet backbone or “cloud”.

Abstract: A system for facilitating image finding analysis includes one or more processors and one or more hardware storage devices storing instructions that are executable by the one or more processors to configure the system to perform acts such as (i) presenting an image on a user interface, the image being one of a plurality of images provided on the user interface in a navigable format, (ii) obtaining a voice annotation for the image, the voice annotation being based on a voice signal of a user, and (iii) binding the voice annotation to at least one aspect of the image, wherein the binding modifies metadata of the image based on the voice annotation.

Abstract: A method of image processing in a structured light imaging system is provided that includes receiving a captured image of a scene, wherein the captured image is captured by a camera of a projector-camera pair, and wherein the captured image includes a binary pattern projected into the scene by the projector, applying a filter to the rectified captured image to generate a local threshold image, wherein the local threshold image includes a local threshold value for each pixel in the rectified captured image, and extracting a binary image from the rectified captured image wherein a value of each location in the binary image is determined based on a comparison of a value of a pixel in a corresponding location in the rectified captured image to a local threshold value in a corresponding location in the local threshold image.

Abstract: Systems and methods are disclosed for generating a specialized machine learning model by receiving a generalized machine learning model generated by processing a plurality of first training images to predict at least one cancer characteristic, receiving a plurality of second training images, the first training images and the second training images include images of tissue specimens and/or images algorithmically generated to replicate tissue specimens, receiving a plurality of target specialized attributes related to a respective second training image of the plurality of second training images, generating a specialized machine learning model by modifying the generalized machine learning model based on the plurality of second training images and the target specialized attributes, receiving a target image corresponding to a target specimen, applying the specialized machine learning model to the target image to determine at least one characteristic of the target image, and outputting the characteristic of the tar