Abstract: A method and device for calculating a river surface flow velocity are provided based on a variational principle, which are used to capture and process the images of an objective area, and to obtain the flow velocity field data of the objective area with high precision in a non-contact manner. The method and device include 3 steps: (1) preparation before initial flow measurement; (2) capturing a video of the river by an image acquisition device, converting a motion of a pixel flow field of the fluid in a captured image sequence into solving an energy functional optimization problem, and solving partial differential equations to obtain data of pixel flow field distribution; and (3) obtaining space coordinates of the pixel point in a world coordinate system and calculating the flow velocity according to the data obtained in the step 2 and the transformation relationship determined in the step 1.

Abstract: The present invention relates to a method and system for automatic localisation of static objects in an urban environment. More particularly, the present invention relates to the use of noisy 2-Dimensional (2D) image data to identify and determine 3-Dimensional (3D) positions of objects in large scale urban or city environments. Aspects and/or embodiments seek to provide a method, system, and vehicle for automatically locating static 3D objects in urban environments by using a voting-based triangulation technique. Aspects and/or embodiments also provide a method for updating map data after automatically new 3D static objects in an environment.

Abstract: A method for pre-operative orthopedic planning includes obtaining only a high-resolution knee-joint scan of a patient, determining hip rotation center and ankle rotation center from anthropometric data based on personal data of the patient, and determining a mechanical axis of the knee joint based on the anthropometric data. The method also includes preparing at least a two-dimensional image model of the knee joint using the knee-joint scan and the determined mechanical axis, and preparing a pre-operative surgical plan based on the image of the knee joint.

Abstract: A computer implemented method for calibrating a camera comprises the following steps carried out by computer hardware components: activating a subset of a plurality of light sources according to a plurality of activation schemes, wherein each activation scheme indicates which of the plurality of light sources to activate; capturing an image for each activation scheme using the camera; and calibrating the camera based on the captured images.

Abstract: A system and method for providing image guidance for placement of one or more medical devices at a target location. The system can determine one or more intersections between a medical device and an image region based at least in part on first emplacement data and second emplacement data. Using the determined intersections, the system can cause one or more displays to display perspective views of image guidance cues, including an intersection indicator in a virtual 3D space.

Abstract: The present invention relates to systems and methods for sequential operation of staining, imaging and sectioning of tissue samples by a processing system. After each layer of the sample is removed by the sectioning system, the system automatically stains the exposed surface of a sample to a depth to enable imaging of the remaining tissue. The system then repeats the sectioning, staining and imaging steps in sequence to image the sample.

Abstract: The disclosure pertains to techniques for image processing. One such technique comprises a method for image selection, comprising: obtaining a sequence of images, detecting a first face in one or more images of the sequence of images, determining a first location for the detected first face in each of the images having the detected first face, generating a heat map based on the first location of the detected first face in each of the images of the sequences of images, determining a face quality score for the detected first face for each of the one or more images having the detected first face, determining a peak face quality score for the detected first face based in part on the face quality score and the generated heat map, and selecting a first image of the sequence of images, corresponding with the peak face quality score for the detected first face.

Abstract: A device determines positions of objects in a scene. The device obtains object detection data (ODD) which identifies the objects and locations of reference points of the objects in 2D images of the scene. The device processes the ODD to generate candidate association data (CAD) which associates pairs of objects between the images, computes estimated 3D positions in the scene for associated pairs of objects in the CAD, and performs clustering of the estimated positions. The device further generates, based on estimated 3D positions in one or more clusters, final association data (FAD) which associates one or more objects between the images, and computes one or more final 3D positions in the scene for one or more reference points of the one or more objects in the FAD. The final 3D position(s) represent the 3D position or the 3D pose of the respective object in the scene.

Abstract: A semi-supervised learning method for object detection in an autonomous vehicle and a device for performing semi-supervised learning for object detection in an autonomous vehicle can include receiving, by a server, no-label voxel data from a vehicle, performing, by the server, a data-based update on a server object detection model on the basis of label voxel data and the no-label voxel data, determining, by the server, a loss value on the basis of the label voxel data and the no-label voxel data, and performing, by the server, a loss-based update on the server object detection model using the loss value.

Abstract: The present disclosure is directed to a method and apparatus for locating a target location on a reaction cell and using the target location to perform an assay. In an example embodiment, a method of performing at least one assay includes obtaining at least one image of a fluid sample located on a reaction cell and creating a set of derivative data including a plurality of derivative data points based on the at least one image. The method also includes determining an image gradient data point for each of the plurality of derivative data points and determining a target location of the fluid sample in the reaction cell based on the image gradient data points. The method further includes performing at least one assay using the target location of the fluid sample in the reaction cell.

Abstract: The present teaching relates to method, system, and programming for responding to an image related query. Information related to each of a plurality of images is received, wherein the information represents concepts co-existing in the image. Visual semantics for each of the plurality of images are created based on the information related thereto. Representations of scenes of the plurality of images are obtained via machine learning, based on the visual semantics of the plurality of images, wherein the representations capture concepts associated with the scenes.

Abstract: A video content matching system includes a computing platform having a hardware processor and a memory storing a software code. When executed, the software code obtains a reference digital profile of a reference video segment, obtains a target digital profile of target video content, and compares the reference and target digital profiles to detect a candidate video segment of the target video content for matching to the reference video segment. The software code also frame aligns reference video frames of the reference video segment with corresponding candidate video frames of the candidate video segment to provide frame aligned video frame pairs, pixel aligns the frame aligned video frame pairs to produce frame and pixel aligned video frame pairs, and identifies, using the frame and pixel aligned video frame pairs, the candidate video segment as a matching video segment or a non-matching video segment for the reference video segment.

Abstract: Systems and methods for selective reporting of construction errors are provided. For example, image data captured from a construction site using at least one image sensor may be obtained. The image data may be analyzed to identify a construction error. The image data may be analyzed to identify a degree of the construction error. The identified degree of the construction error may be compared with a threshold selected based on an analysis of a construction plan associated with the construction site. In response to a first result of the comparison, a first severity may be assigned to the construction error, and in response to a second result of the comparison, a second severity may be assigned to the construction error, the second severity differs from the first severity. Further, information related to the construction error may be provided based on the severity assigned to the construction error.

Abstract: In one embodiment, a system identifies a road to be navigated by an ADV, the road being captured by one or more point clouds from one or more LIDAR sensors. The system extracts road marking information of the identified road from the point clouds, the road marking information describing one or more road markings of the identified road. The system partitions the road into one or more road partitions based on the road markings. The system generates a point cloud map based on the road partitions, where the point cloud map is utilized to perceive a driving environment surrounding the ADV.

Abstract: A neural network receives mono-color pixels in a Bayer pattern from an image sensor and interpolates pixels to generate full-color RGB pixels while also enlightening the image for better detail. A Back-Projection (BP) layer is added to each contracting layer in a U-net convolution neural network where feature depth is increased as pixels are downsampled, while a BP channel shrink layer is added to each expansion layer where feature depth is reduced to upsample and increase pixel resolution. Convolution layers in the BP layer lighten and darken images to generate an error that is then lightened and added to a weighted lightened input. The neural network learns the best weightings to correct for noise and error in the lightening process in these BP layers. Noise is reduced further by repeating a convolution and leaky Rectified Linear Unit (lrelu) in series for the first convolution in the BP layer that performs lightening.

Abstract: In part, the disclosure relates to method for identifying regions of interest in a blood vessel. The method includes the steps of: providing OCT image data of the blood vessel; applying a plurality of different edge detection filters to the OCT image data to generate a filter response for each edge detection filter; identifying in each edge detection filter response any response maxima; combining the response maxima for each edge detection filter response while maintaining the spatial relationship of the response maxima, to thereby create edge filtered OCT data; and analyzing the edge filtered OCT data to identify a region of interest, the region of interest defined as a local cluster of response maxima. In one embodiment, one or more indicia are positioned in one or more panels to emphasize a reference vessel profile as part of a user interface.

Abstract: A system and method for recognizing an object. The system includes an imaging apparatus for capturing an image of an object and a processor for receiving the captured image of the object, and for, when it is determined that fewer than a predetermined number of objects have been previously imaged, for determining whether the image of the captured object includes one or more characteristics determined to be similar to a same characteristic in a group of previously imaged objects so that the captured image is grouped with the previously imaged objects, or whether the image of the captured object includes one or more characteristics determined to be dissimilar to a same characteristic in a group of previously imaged objects so that the captured image is not grouped with the previously imaged objects, and the image of the captured object starts another group of previously imaged objects.

Abstract: A method of building an abnormality quantifier comprising: generating at least one selected first dataset comprising measurements of a normal population or sample and at least one second selected dataset comprising measurements of an abnormal population or sample; generating an image or map by imagizing the datasets; identifying a normality zone within the image or map using the first dataset; identifying an abnormality zone within the image or map using the second dataset; determining a definition of abnormality based on a comparison of the normality zone and the abnormality zone; receiving or accessing at least one third dataset comprising measurements of a both known normal and abnormal population or sample; testing the performance of the initially defined abnormality against one or more preset performance criteria; and outputting an abnormality quantifier when optimal performance has been reached.

Abstract: A model training method and an electronic device are provided. The method includes the following steps: establishing a brain age prediction model according to a training set; adjusting a parameter in the brain age prediction model according to a validation set; inputting a test set into the brain age prediction model with the adjusted parameter to obtain a plurality of first predicted brain ages; determining whether the first predicted brain ages satisfy a first specific condition; and completing training of the brain age prediction model when the first predicted brain ages satisfy the first specific condition.

Abstract: One variation of method includes: serving setup frames to a projector facing a scene; at a peripheral control module comprising a camera facing the scene, recording a set of images during projection of corresponding setup frames onto the scene by the projector and a baseline image depicting the scene in the field of view of the camera; calculating a pixel correspondence map based on the set of images and the setup frames; transforming the baseline image into a corrected color image—depicting the scene in the field of view of the camera—based on the pixel correspondence map; linking visual assets to discrete regions in the corrected color image; generating augmented reality frames depicting the visual assets aligned with these discrete regions; and serving the augmented reality frames to the projector to cast depictions of the visual assets onto surfaces, in the scene, corresponding to these discrete regions.