Abstract: A method for transferring data from a medical device to a server comprises receiving a video stream from the medical device, capturing an image from the video stream, transmitting the image to the server via a data network, and extracting the data from the image. The image may illustrate and/or represent data over a period of time. The method may also comprise transmitting, from a data module receiving the video stream from the medical device, a signal to a router that indicates that the data module is connected to the network. The method may also comprise transmitting a command to the data module to start capturing the image, transferring the image to the router, broadcasting a signal indicating that the data module has captured the image, receiving the broadcasted signal at the server, and storing the image at the server.

Abstract: In one embodiment, a method includes accessing an input image associated with an initial illumination, extracting from the input image at least a first, a second, and a third mono-channel images based on a first, a second, a third image channel, respectively, determining a first, a second, and a third correction for the first, second, and third mono-channel images with a first, a second, and a third correction algorithm, respectively, generating a first, a second, and a third corrected mono-channel image by applying the first, second, and third correction to the first, second, and third mono-channel image, respectively, and generating an output corrected image corresponding to the input image based on the first, second, and third mono-channel corrected images, wherein the output corrected image is associated with a corrected illumination with respect to the initial illumination.

Abstract: A vision-based wafer pre-alignment platform and alignment method perform image acquisition and processing based on two groups of binocular cameras. To maintain an outer edge contour of a wafer, an adaptive image denoising method of a wafer image is proposed pertinently. Further, the present disclosure defines edge feature points of the wafer according to shape features of the wafer to form feature points constituting a triangle, facilitating extracting the feature points according to an AAM model and matching the feature points according to an isosceles triangle rule. Finally, a position deviation of the wafer is calculated according to spatial coordinates of the feature points and positions of the feature points, an alignment deviation e is calculated according to the obtained spatial position coordinates of the feature points, and the wafer is pre-aligned based on the deviation value e.

Abstract: Provided is a training method for an image recognition model, a spinneret plate detection method, a spinneret plate detection device, and a storage medium, relating to the fields of image recognition and deep learning technologies. The training method includes: processing a spinneret plate sample image based on a first image recognition model to obtain annotation state information of each micro hole in the spinneret plate sample image, wherein the spinneret plate sample image is a projection image formed on an imaging member after light emitted by a light source passes through each micro hole of a spinneret plate sample and is amplified by an amplifying member, the annotation state information is used to characterize a shape state of a micro hole corresponding thereto; and training a preset image recognition model based on the spinneret plate sample image and the annotation state information to obtain a target image recognition model.

Abstract: Disclosed herein are system, method, and computer program product embodiments for detecting a potential damage caused to a vehicle when the vehicle is placed on a transporter. The method includes obtaining a first image and a second image taken at different time instances. The first image and the second image are generated by one or more cameras placed around a vehicle on a transporter, and cover a same portion of a body of the vehicle. The method further includes detecting a change between a portion of the first image corresponding to the portion of the body of the vehicle and a portion of the second image corresponding to the portion of the body of the vehicle indicative of a collision. Moreover, the method includes flagging a potential damage to the portion of the body of the vehicle based on the detected change between the first image and the second image.

Abstract: A speech video generation device according to an embodiment includes a first encoder that receives an input of a first person background image of a predetermined person partially hidden by a first mask, and extracts a first image feature vector from the first person background image, a second encoder, which receives an input of a second person background image of the person partially hidden by a second mask, and extracts a second image feature vector from the second person background image, a third encoder, which receives an input of a speech audio signal of the person, and extracts a voice feature vector from the speech audio signal, a combining unit, which generates a combined vector of the first image feature vector, the second image feature vector, and the voice feature vector, and a decoder, which reconstructs a speech video of the person using the combined vector as an input.

Abstract: A medical image processing apparatus includes processing circuitry configured to acquire medical image data representing a region including a target site, divide the medical image data into a plurality of pieces of divisional data according to an anatomical structure, extract a target region corresponding to the target site from each of the pieces of divisional data according to a particular condition, and cause an output circuit to output information on the extracted target region.

Abstract: An example method includes receiving images that each include an optical code. The method further includes cropping the images to isolate the optical codes and rotating the images to achieve a desired orientation for the optical code of each of the images. The method further includes rectifying, after the cropping and the rotating, a perspective of each of the images to achieve a desired perspective angle for the optical code of each of the images. The method further includes deblurring, after the rectifying, the images to reduce blurring in each of the images. The method further includes binarizing, after the deblurring, the images to correct pixels of the optical code in each of the images. The method further includes performing, after the binarizing, an optical code reading process on the plurality of images to decode optical codes in the images.

Abstract: Provided herein are textile items return systems including which are configured to identify the type and/or condition of a returned textile item. The systems include an image acquisition device configured to acquire image data of a returned textile item and identification module which is configured to receive the acquired image data and identify one or more item parameters of the returned textile item using a suitable classification algorithm. Further provided are textile items dispensing and return (TIDR) systems including the textile items return system.

Abstract: A three-dimensional object detecting device generates a mask image that masks regions outside a three-dimensional object candidate region in a difference image of a first overhead image and a second overhead view image for which the imaging locations O are mutually aligned, identifies a near ground contact line of a three-dimensional object based on a masked difference image. The difference image is masked with the mask image, finds an end point of the three-dimensional object based on the masked difference image, identifies the width of the three-dimensional object based on a distance between a non-masking region boundary and the end point of the three-dimensional object in the mask image, identifies a far ground contact line of the three-dimensional object based on the width of the three-dimensional object and the near ground contact line, and identifies the location of the three-dimensional object in the difference image based on the near ground contact line and the far ground contact line.

Abstract: The present invention proposes a technique for enabling the execution of measurement processing without referring to a design drawing for which it is difficult to adjust or obtain parameters for image processing that requires knowhow. This measurement system according to the present disclosure refers to a learning model generated on the basis of teaching data, which is generated from a sample image of a semiconductor, and the sample image, generates a region-segmented image from an input image (measurement subject) of a semiconductor having a predetermined structure, and uses the region-segmented image to perform image measurement. Here, the teaching data is an image in which labels, which include a structure of the semiconductor in the sample image, are assigned to each pixel of the image, and the learning model includes parameters for deducing teaching data from the sample image (see indicator 1).

Abstract: Systems, methods, and software of inspecting growth areas for plants. In one embodiment, an inspection system captures a plurality of digital images of the growth area from different angles in relation to the growth area, and processes the digital images to identify a boundary of the growth area in the digital images. The inspection system combines the digital images based on the boundary identified in the digital images to generate a composite image of the growth area, and performs image processing on the composite image to detect one or more deficient growth regions in the growth area. The inspection system highlights the deficient growth regions in the composite image, and displays the composite image with the deficient growth regions highlighted.

Abstract: A method and apparatus for analyzing diagnostic image data are provided in which a plurality of acquisition images of a vessel of interest having been acquired with a pre-defined acquisition method is received at a trained classifying device and classified, by the classifying device, to extract at least one quantitative feature of the vessel of interest from at least one acquisition image of the plurality of acquisition images. The at least one quantitative feature is then output associated with the at least one acquisition image while the acquisition of the diagnostic image data is still in progress and one or more adjustable image acquisition settings are adjusted based on the at least one quantitative feature to optimize the acquisition of the diagnostic image data.

Abstract: A system and method for real time discrete cosine transform image and video processing with convolutional neural network architecture. The system and method incorporate discrete cosine transform image processing with convolutional neural networks to achieve fast and efficient image processing that yields more reliable results than previously used image processing methods. The proposed system and method enable effective, real time, image processing which is applicable to a wide range of imaging and video devices.

Abstract: Face detection in a spherical image is performed using overcapture. Multiple views of a spherical image are separately processed using face detection, and results of the face detection for those views are projected to a format for further processing of the spherical image.

Abstract: A non-contact facial blood pressure measurement method based on 3D CNN is disclosed, which belongs to the technical field of computer vision. The method includes the following steps. S110: collecting an actual face video sample and training a blood pressure prediction model based on face images using 3D CNN neural network. S120: obtaining a face video in real time through a HD camera. S130: recognizing face key points in the face video obtained in S120 through dlib face recognition model, selecting a face region of interest, and extracting face images from the region. S140: performing a wavelet transform operation on the face images extracted in S130 to remove noise. S150: inputting seven consecutive frames of the face images into the 3D CNN blood pressure prediction model trained in S110 to obtain a blood pressure value of the measured person. The disclosure realizes non-contact facial blood pressure measurement.

Abstract: Method and systems are described that create a 3D reconstruction of a vessel of interest that represents a subset of a coronary tree that includes a lesion; calculate at least one of pressure parameters or anatomical parameters based at least in part on a portion of the 3D reconstruction that includes the lesion; calculate a wall shear stress (WSS) descriptor, based on the 3D reconstruction, for a segment of a surface of the vessel that includes the lesion, wherein the WSS descriptor includes information regarding an amount of variation in contraction or expansion applied at surface elements within the segment during at least a portion of a cardiac cycle; and calculate a myocardial infarction (MI) index based on the WSS descriptor and the at least one of the pressure or anatomical parameters, the MI index representing a likelihood that the lesion will result in an MI.

Abstract: A method for controlling a warning system including a user device. The method includes, in the warning system: receiving a wireless signal transmitted towards a surface on which a user is located, of which the range defines, on the surface, a plurality of zones including a no-warning first zone and a warning second zone, a characteristic relating to the transmitted signal being able to take a second value corresponding to the second zone; upon detecting the characteristic having the second value, generating a warning signal, and transmitting the warning signal to a control unit of the warning system in order to notify a user of the system.

Abstract: A method for determining images plausible to have a false negative object detection comprises providing a group of historic trajectories, wherein each historic trajectory comprises a reference track that represents one or more historic tracks and comprises an object class of historic object detections that belong to the one or more historic tracks; performing tracking; performing object detection; for a determined track that does not match any determined object detection, comparing the determined track with reference tracks of historic trajectories for identifying a matching reference track; upon identifying a matching reference track, defining images of the determined track as being plausible to have a false negative object detection for the object class of the historic trajectory comprising the matching reference track; and upon not identifying a matching reference track, defining the determined track as a false positive track.

Abstract: A method for evaluating images of a printed pattern. The method includes obtaining a first averaged image of the printed pattern, where the first averaged image is generated by averaging raw images of the printed pattern. The method also includes identifying one or more features of the first averaged image. The method further includes evaluating the first averaged image, using an image quality classification model and based at least on the one or more features. The evaluating includes determining, by the image quality classification model, whether the first averaged image satisfies a metric.