Disclosed are systems, devices, and methods for detecting a trachea, an exemplary system comprising an imaging device configured to obtain image data and a computing device configured to generate a three-dimensional (3D) model, identify a potential connected component in a first slice image, identify a potential connected component in a second slice image, label the first slice image as a top slice image, label the connected component in the top slice image as an active object, associate each connected component in a current slice image with a corresponding connected component in a previous slice image based on a connectivity criterion, label each connected component in the current slice image associated with a connected component of the preceding slice image as the active object, and identify the active object as the trachea, based on a length of the active object.

A method for processing video data comprising a plurality of image frames, the method comprising: obtaining a forward model and a backward model of the plurality of image frames; processing the forward model and the backward model to propagate at least one label in the region or patch to at least one other image frame of the video sequence, using a probabilistic method for estimating the label in the at least one other image frame in forward and backward correspondences, wherein, during the processing, a pixel having a most likely label with a probability lower than a threshold value is assigned a predetermined generic label; and generating a labelled result for any given image frame by applying an image label difference, based on label uncertainty between the forward and backward correspondences, to the given image frame.

An image processing device includes an edge detector to detect a portion in which a luminance difference between neighboring pixels changes in a direction that increases equal to or more than a first predetermined value as a positive edge and a portion in which the luminance difference between neighboring pixels changes in a direction that decreases equal to or more than a second predetermined value as a negative edge, a grouping portion configured to group the positive edges that have been detected into a positive edge group and the negative edges that have been detected into a negative edge group, and a boundary line setting portion configured to set, as a boundary line: (i) a line connecting end points between the positive edges in the positive edge group; or (ii) a line connecting end points between the negative edges in the negative edge group.

A device that includes a model training engine implemented by a processor. The model training engine is configured to obtain a set of data values associated with a feature vector. The model training engine is further configured to generate a set of gradients by dividing separation distances by an average separation distance and to compare each gradient to a gradient threshold value. The model training engine is further configured to identify a boundary in response to determining a gradient exceeds the gradient threshold value, to determine a number of identified boundaries, and to determine a number of clusters based on the number of identified boundaries. The model training engine is further configured to train the machine learning model to associate the determined number of clusters with the feature vector.

A technique facilitates construction of a wire-wrapped screen. A wrapping machine is operated with a sensor, e.g. a camera, positioned adjacent the wrapping machine while wire is wrapped to create the wire-wrapped screen. The sensor is used to obtain data on at least one parameter of the wire-wrapped screen during creation of the wire-wrapped screen. Data is provided to a controller in communication with the wrapping machine to improve the quality of the wire-wrapped screen. For example, data from the images obtained via the camera may be provided to the controller which is configured to determine slot width as the wire is wrapped. The controller is then able to provide feedback in real time to the wrapping machine so as to adjust the wrapping machine for maintaining a desired slot width.

A computer implemented method of feature detection in temporal graph data structures of events, the method including receiving a temporal series of graph data structures of events each including a plurality of nodes corresponding to events and edges connecting nodes corresponding to relationships between events; rendering each graph data structure in the series as an image representation of the graph data structure including a representation of nodes and edges in the graph being rendered reproducibly in a cartesian space based on attributes of the nodes and edges, so as to generate a temporal series of image representations ordered according to the temporal graph data structures; processing the series of image representations by a convolutional neural network to classify the image series so as to identify a feature in the image series, the convolutional neural network being trained by a supervised training method including a plurality of training example image series in which a subset of the training examples are classified as including the feature.

The present disclosure relates to verifying an initial object estimation of an object. A two-dimensional (2D) image representative of an environment including one or more objects may be obtained. The 2D image may be inspected to detect edges of an object. The edges may be processed to verify or update an initial object estimation to increase the accuracy of an object detection result.

In various examples, operations include obtaining, from a machine learning model, feature classifications that correspond to features of objects depicted in images of a geographical area in which the images are provided to the machine learning model. The operations may also include annotating the images with three-dimensional representations that are based on the obtained feature classifications. Further, the operations may include generating map data corresponding to the geographical area based on the annotated images.

The present disclosure provides a computer-implemented method, a device, and a computer program product for deep lesion tracker. The method includes inputting a search image into a first three-dimensional DenseFPN (feature pyramid network) of an image encoder and inputting a template image into a second three-dimensional DenseFPN of the image encoder to extract image features; encoding anatomy signals of the search image and the template image as Gaussian heatmaps, and inputting the Gaussian heatmap of the template image into a first anatomy signal encoders (ASE) and inputting the Gaussian heatmap of the search image into a second ASE to extract anatomy features; inputting the image features and the anatomy features into a fast cross-correlation layer to generate correspondence maps, and computing a probability map according to the correspondence maps; and performing supervised learning or self-supervised learning to predict a lesion center in the search image.

An approach is provided for detecting and map-coding a tunnel based on probes and image data. The approach involves, for example, identifying a gap in probe data collected from one or more location sensors of a plurality vehicles. The gap represents a probe gap segment along which at least one probe point of the probe data does not occur or occurs below a threshold number. The approach also involves retrieving image data depicting a geographic area based on location coordinate data associated with the gap. The approach further involves processing the image data to identify one or more end points of a road network depicted in the image data. The approach further involves locating a tunnel start point and/or a tunnel end point based on the one or more endpoints. The approach further involves providing the tunnel start point and/or the tunnel end point as a map data output.