Abstract: An example system to capture digital images includes: a colposcope device configured to capture a digital image during a colposcopy procedure; and an image processing module programmed to digitally process the digital image, including: a filter module programmed to filter certain aspects of the digital image; and an overlay module programmed to allow the digital image to be annotated. In other examples, colposcope devices include: a cradle, the cradle being configured to hold a portable computing device, the cradle including an aperture; and an optical capture device, the optical capture device including a constant magnification lens system, the constant magnification lens system being configured to direct light through the aperture.

Abstract: This disclosure describes methods and systems for improving performance of biometric systems that use features in the eye, such as iris or eye-veins, particularly when both biometric performance and user convenience are objectives. The disclosure relates to optimizing biometric performance when the iris or eye veins are neither fully visible, nor fully occluded, but in a partially-occluded state which occurs often when a user's eye is in a relaxed or natural state. In some embodiments, the method comprises a biometric enrollment or training step whereby an original image of a human eye is acquired, and a plurality of synthetic or augmented images are generated that are a combination of the original image and synthesized images that simulate specific ways that the eye can be occluded. A classifier can be trained using the plurality of augmented reference images, and subsequent recognition is performed using the classifier on newly acquired real images.

Abstract: A method and system for generating an output image from a plurality, N, of corresponding input image channels is described. A Jacobian matrix of the plurality of corresponding input image channels is determined. The principal characteristic vector of the outer product of the Jacobian matrix is calculated. The sign associated with the principal characteristic vector is set whereby an input image channel pixel projected by the principal characteristic vector results in a positive scalar value. The output image as a per-pixel projection of the input channels in the direction of the principal characteristic vector is generated.

Abstract: In the current artificial intelligence field, models of deep learning that is prevalent can only map functions. Therefore, a machine learning model with higher performance is desirable. The issue is to construct a machine learning model that enables deep competitive learning between data based on the exact distance. A precise distance scale is submitted by unifying Euclidean space and probability space. It submits a measure of the probability measure of fuzzy event based on this distance. Or, it constructs a new neural network that can transmit information of the maximum probability. Furthermore, super deep competition learning is performed between data having very small ambiguous fuzzy information and minute unstable probability information. By performing integral calculation on this result, it has become possible to obtain dramatic effects at tape macro level.

Abstract: Among other aspects, various embodiments include encoding wavelength-based characteristics, in addition to three-dimensional positions, of a plurality of objects of a plurality of different wavelengths directly in an image of the objects.

Abstract: Provided is a specific object detection apparatus capable of detecting a specific object even when a self-vehicle is positioned at a distance from the specific object during bad weather.

Abstract: Embodiments of the disclosure provide systems and methods for segmenting a medical image. An exemplary system includes a communication interface configured to receive the medical image acquired by an image acquisition device. The system further includes a memory configured to store a multi-level learning network including at least a first convolution block and a second convolution block. The second convolution block has at least one convolution layer. The system also includes a processor. The processor is configured to determine a first feature map by applying the first convolution block to the medical image, and determine a second feature map by applying the second convolution block to the first feature map. The processor is further configured to determine a first level feature map by concatenating the first feature map and the second feature map. The processor is also configured to obtain a first level segmented image based on the first level feature map.

Abstract: A method is provided for measuring or estimating stress distributions on heart valve leaflets by obtaining three-dimensional images of the heart valve leaflets, segmenting the heart valve leaflets in the three-dimensional images by capturing locally varying thicknesses of the heart valve leaflets in three-dimensional image data to generate an image-derived patient-specific model of the heart valve leaflets, and applying the image-derived patient-specific model of the heart valve leaflets to a finite element analysis (FEA) algorithm to estimate stresses on the heart valve leaflets. The images of the heart valve leaflets may be obtained using real-time 3D transesophageal echocardiography (rt-3DTEE). Volumetric images of the mitral valve at mid-systole may be analyzed by user-initialized segmentation and 3D deformable modeling with continuous medial representation to obtain, a compact representation of shape.

Abstract: The present invention relates to systems and methods for adaptively optimizing broadband reference spectra for a multi-spectral image or adaptively optimizing reference colors for a bright-field image. The methods and systems of the present invention involve optimization techniques that are based on structures detected in an unmixed channel of the image, and involves detecting and segmenting structures from a channel, updating a reference matrix with signals estimated from the structures, subsequently unmixing the image using the updated reference matrix, and iteratively repeating the process until an optimized reference matrix is achieved.

Abstract: Disclosed are a method and a system for recognizing a license plate. The method for recognizing a license plate according to an embodiment of the present disclosure includes generating a localization model for localizing a license plate area attached to an object by learning using image data including the preset object, and a recognition model for recognizing identification information of the license plate by learning using preset license plate image data; when input image is received, localizing a license plate area from the input image using the localization model; and recognizing identification information of a license plate of the corresponding object from the localized license plate area using the recognition model.

Abstract: An approach is provided for an asymmetric evaluation of polygon similarity. The approach, for instance, involves receiving a first polygon representing an object depicted in an image. The approach also involves generating a transformation of the image comprising image elements whose values are based on a respective distance that each image element is from a nearest image element located on a first boundary of the first polygon. The approach further involves determining a subset of the plurality of image elements of the transformation that intersect with a second boundary of a second polygon. The approach further involves calculating a polygon similarity of the second polygon with respect the first polygon based on the values of the subset of image elements normalized to a length of the second boundary of the second polygon.

Abstract: A tracking and ranging system includes a thermal sensor device, a controller, a ranging device and a transmission device. The thermal sensor device is configured to capture a thermal image. The controller analyzes the thermal image to identify the main heat source from among the heat sources displayed in the thermal image, and obtain an offset distance between the center points of the main heat source and the thermal image. The ranging device is coupled to the controller. The transmission device loads the ranging device and is coupled to the controller. The controller controls the motion of the transmission device in accordance with the offset distance to correct the offset angle between the ranging device and the object corresponding to the main heat source. After correcting the offset angle, the ranging device detects a first distance to the object by transmitting energy and receiving reflected energy.

Abstract: Dynamic methods and systems for monitoring an environment. The methods and systems adopt functional nodes with at least one context node and one or more application nodes. The application nodes include at least one of a processing application node, a notifying application node, and/or an action application node. Input data of the environment are analyzed and detected. Representative data are extracted and compared with a group of comparative data by calculating a deviation between the representative data and the most similar comparative data. The notifying application node may send an alarm as a function of the deviation detected as above. An environment monitoring network is formed as a function of the functional nodes. The environment monitoring network is topologically variable as a function of i) the environment to be monitored, ii) the functions of the functional nodes and iii) the deviation detected. Monitoring occurs as a function of the topology of the implemented environment monitoring network.

Abstract: A system for accurately determining a size of a pet includes receiving images of the pet (at least three images—front, side, and top are preferred). The system utilizes software (e.g. Artificial Intelligence) to determine a breed of the pet and base sizes of the pet, then the software further hones the base sizes by analysis of the images and, in some embodiments, data from returned merchandise. Once the size is determined, products are presented to the consumer that are in stock and available in the size of the pet. In some embodiments, images of the pet are overlaid with at least one of the products and displayed showing the product on the pet.

Abstract: This disclosure describes a system for automatically identifying an item from among a variation of items of a same type. For example, an image may be processed and resulting item image information compared with stored item image information to determine a type of item represented in the image. If the matching stored item image information is part of a cluster, the item image information may then be compared with distinctive features associated with stored item image information of the cluster to determine the variation of the item represented in the received image.

Abstract: For omnidirectional video such as 360-degree Virtual Reality (360VR) video, a video system that support independent decoding of different views of the omnidirectional video is provided. A decoder for such a system can extract a specified part of a bitstream to decode a desired perspective/face/view of an omnidirectional image without decoding the entire image while suffering minimal or no loss in coding efficiency.

Abstract: Exemplary systems and methods perform post-processing operations on computer vision model detections of objects of interest in geospatial imagery to detect and assign attributes to the detected objects of interest. For example, an exemplary post-processing system correlates multiple detections, made by a computer vision model, of an object of interest depicted within a set of images of a geospatial location, determines, based on the correlated detections, an attribute of the object of interest depicted within the set of images of the geospatial location, and selects the attribute for inclusion in a dataset for the object of interest. Corresponding methods and systems are also disclosed.

Abstract: Systems and methods are provided for to estimate the pose of a human subject's head from a sequence of images received from a single depth camera by processing the images to generate a continuous estimate of the head pose in a 3-dimensional (3D) space, and to generate a 3D head model for display and further use. The subject is provided instructions to rotate their head in a first direction until a threshold angle of rotation is reached and then are provided instructions to rotate their head in a second direction. The depth camera provides a sequence of captured images which are processed to extract head meshes. After capture is complete the head meshes are merged to generate a 3D model of the subject's head.

Abstract: A visual indication is generated of depth locations of slices obtained by applying a reconstruction algorithm in a computed tomography method with given settings, the display being shown on the object, e.g. on the patient's lateral side or in close relation to the patient.

Abstract: A method for digital subtraction angiography of a recording region of a patient is provided herein. The method includes recording at least one mask image of a recording region without using a contrast medium; recording a plurality of ill images after administration of a contrast medium; and determining result images by subtraction of one of the at least one mask image from respective fill images.