Abstract: A bottle storage includes a housing portion that houses a bottle therein, a cooling-warming portion that is provided around the housing portion, a sensor which detects insertion of the bottle into the housing portion, an image capture portion that captures an image of the bottle housed in the housing portion, a controller that causes the image capture portion to perform image capture for capturing the image of the bottle when the bottle is sensed to be inserted into the housing portion by the sensor, an information manager that identifies at least a brand of the bottle, based on recognition result of the bottle image captured by the control of the controller, and an output portion that outputs characteristics of a beverage of the brand identified by the information manager.

Abstract: A protein crystallization trial is automatically analyzed by capturing images of the protein drops in the trial. A machine-learned model, such as a neural network, is applied to classify the images. The model generates a predicted classification from among a set of possible classifications which includes one or more crystal type classifications and one or more non-crystal type classifications. Users may be notified automatically of newly identified crystals (e.g., drops that are classified as a crystal type). The notification may include a link to a user interface that includes results of the trial.

Abstract: Methods and systems for PET image reconstruction are provided. A method may include obtaining an image sequence associated with a subject. The image sequence may include one or more images generated via scanning the subject at one or more consecutive time periods. The method may also include obtaining a target machine learning model. The method may further include generating at least one target image using the target machine learning model based on the image sequence. The at least one target image may present a dynamic parameter associated with the subject. The target machine learning model may provide a mapping between the image sequence and the at least one target image.

Abstract: A digital pathology system and associated method and computer program product provide a quantitative analysis of entire tissue slides as well as intuitive, effective, fast, and precise quantification of biomarker expressions across relevant areas of the entire tissue slides. The digital pathology system enables a novel workflow that allows the user to efficiently outline clinically relevant morphology in its entirety, including solid tumor areas. Quantitative analysis results are then efficiently and intuitively provided to the user for all tissue content (i.e., millions of cells) within seconds. This efficiency is made possible by a pre-computation step that computes and stores image analysis results for later retrieval. Visualizing vast amount of data effectively is another component of the system that provides information and confidence to the user about the biomarker expression levels.

Abstract: Various systems and methods for generating images are provided. In some embodiments, the techniques can include acquiring a medical image and an associated motion characterization. The motion characterization can then be used to generate a plurality of gated image data sets, sorted by phase in the motion cycle. A new amplitude-based motion characterization curve is derived from the association of phases with amplitude-based characteristics in the phase gated images. This newly derived amplitude-based motion characterization curve can then be used to re-sort data according to amplitude-based gating techniques known in the field or with data driven optimization techniques.

Abstract: A face detection device and method are provided. The face detection device receives captured images from a camera, and extracts vital information from a face region included in the captured images. The face detection device determines, based on the vital information, whether or not a face included in the face region is in a living state, and collates a feature quantity of the face included in the face region with information of a face which is previously registered. The face detection device selects, according to a user input, whether or not to perform at least one or both of a process for the determining, and a process for the collating.

Abstract: Sensor data, including a group of time series elements, is received. A training data set is determined, including by determining for at least a selected time series element in the group of time series elements a corresponding ground truth. The corresponding ground truth is based on a plurality of time series elements in the group of time series elements. A processor is used to train a machine learning model using the training dataset.

Abstract: The disclosure relates to a system for automatic recognition of plants, including a central server unit, a network, an automatic picture processing unit, a plant database unit, in which plant data are stored, a picture database unit, in which picture data related to plants are stored, and a mobile device with a camera, the camera of the mobile device configured to provide picture data of a plant and the picture data transmitted via the network to the server unit, the automatic picture processing unit together with the plant database unit and the picture database unit on the basis of the transmitted picture data is configured to perform a comparison with data in the plant database unit and the picture database unit and the transmitted picture data of the plant such, that the plant is defined and the definition can be is transmitted back to the mobile device.

Abstract: An automatically calibrated vehicle-tracking system and methods of use thereof. The automatically calibrated vehicle-tracking system has an input interface for receiving an image stream from a tracking camera and vehicle license plate data indicative of valid license plate detections from a license plate camera; a general purpose processor; a computer-readable memory comprising calibration program code for calibrating the vehicle-tracking system, the calibration program code comprising: a tracking module to generate a plurality of calibration tracks, a pairing module to identify, for each of the plurality of calibration tracks, an association between a valid license plate detection and the calibration track and a calibration to set a threshold for a track parameter.

Abstract: A vehicle capable of autonomous driving includes a lane detection system. The lane detection system is trained to predict lane lines using training images. The training images are automatically processed by a training module of the lane detection system in order to create ground truth data. The ground truth data is used to train the lane detection system to predict lane lines that are occluded in real-time images of roadways. The lane detection system predicts lane lines of a roadway in a real-time image even though the lane lines maybe indiscernible due to objects on the roadway or due to the position of the lane lines being in the horizon.

Abstract: A seam inspection apparatus can automatically determine the quality of an image of a seam of a sewn product. The seam inspection apparatus includes an image data acquisition unit that acquires image data of a seam of a sewn product, a feature extraction unit that extracts a feature quantity of the seam from the image data of the seam of the sewn product acquired by the image data acquisition unit, and a quality determination unit that performs quality determination of the seam based on the feature quantity of the seam.

Abstract: A system for updating a point cloud of an environment having markers for measuring an environment and a method are provided. The system includes a mobile device having a camera. The system further includes one or more processors responsive to executable computer instructions to perform a method comprising capturing one or more images of at least a portion of the environment that include the one or more markers in the one or more images, identifying the one or more markers within the one or more images, comparing the identified one or more markers to a reference system with known coordinates and integrating the one or more images into a point cloud.

Abstract: A system for object recognition and segmentation from digital images provides an intelligent object recognition and segmentation using one or more multilayer convolutional neural network (CNN) models trained in multiple-stages and in a parallel and distributed manner to improve training speed and efficiency. The training dataset used in each of the multiple training stages for the CNN models are generated, expanded, self-validated from a preceding stage. The trained final CNN models are augmented with post-model filters to enhance prediction accuracy by removing false positive object recognition and segmentation. The system provides improved accuracy to predict object labels to append to unlabeled image blocks in digital images. In one embodiment, the system may be useful for enhancing a digital landmark registry by appending identifying labels on new infrastructure improvements recognized in aerial or satellite land images.

Abstract: A method is provided for model training to detect defective products. The method includes sampling training images of a product to (i) extract image portions therefrom made of a center patch and its context and (ii) black-out the center patch. The method further includes performing unsupervised back-propagation training of a Contextual Auto-Encoder (CAE) model using (i) the image portions with the blacked-out center patch as an input and, (ii) the center patch as a target output and, (iii) an image-based loss function, to obtain a trained CAE model. The method also includes sampling positive and negative center-patch-sized portions from the training images. The method additionally includes normalizing, using the trained CAE model, the positive and negative center-patch-sized portions.

Abstract: Embodiments of this application disclose a method for determining camera pose information of a camera of a mobile terminal. The method includes: obtaining a first image, a second image, and a template image; performing feature point detection on a first feature point of the template image and a second feature point of the second image, to obtain a first homography; determining a first target homography according to a first optical flow feature point in the first image and a second optical flow feature point in the second image, and determining a second homography according to the first target homography and a second target homography; and performing complementary filtering processing on the first homography and the second homography, to obtain camera pose information of the camera. In the embodiments of this application, complementary filtering processing may be performed on two homographies obtained in a camera pose tracking process.

Abstract: An image generation device includes a first reception unit and a control unit. The first reception unit receives travel information about the travel state of a moving body. The control unit sets, based on the travel information, a multiple-exposure region in an imaging region of an image sensor used on the moving body, and generates image data in which the multiple-exposure region in the imaging region is formed through a multiple exposure and a region other than the multiple-exposure region is not formed through the multiple exposure.

Abstract: A seam inspection device includes an imaging device and a processing device. The imaging device shoots a sewing object supported by a throat plate of a sewing machine and having seams formed therein. And the processing device detects an abnormality of the seam based on an image of the sewing object acquired by the imaging device.

Abstract: The present invention provides a medical image processing method and a computer-readable storage medium. The method includes: reconstructing a two-dimensional cross-sectional image of an imaged tissue based on a volumetric image of the imaged tissue; projecting a CT value of the imaged tissue along a normal direction of the centerline of the imaged tissue in the two-dimensional cross-sectional image; and, positioning the imaged tissue based on the projection result of the CT value of the imaged tissue.

Abstract: A method of distant on-road object detection, comprising, capturing an initial frame of a roadway lane in a direction of travel, detecting lane edges in the initial frame, capturing a subsequent frame of the roadway lane, cropping the subsequent frame based on the detected lane edges of the initial frame resulting in an image patch, detecting an image patch object within the image patch, down-sampling the subsequent frame, detecting a down-sampled object within the down-sampled subsequent frame and merging the detections within the image patch and the detections within the down-sampled subsequent frame.

Abstract: In various embodiments, an image monitoring device (100) may acquire (200) video data that captures a medical room. Image processing may be performed on the acquired image data to perform the following tasks: identifying (202) at least one medical apparatus (402) in the medical room; identifying (204) a medical personnel in the medical room; detecting (206) that the medical personnel has occluded a line of sight between the image monitoring device and the at least one medical apparatus; and determining (206) an amount of time that the medical personnel occludes the line of sight. In various embodiments, the amount of time may be compared to one or more thresholds set forth by the first medical protocol. An alert may be triggered (208) in response to a determination that the amount of time fails to satisfy a first threshold of the one or more thresholds.