Abstract: An image generation apparatus includes a processing circuit and a memory storing at least one computational image. The at least one computational image is a light-field image, a compressive sensing image, or a coded image. The processing circuit (a1) identifies a position of an object in the at least one computational image using a classification device, (a2) generates, using the at least one computational image, a display image in which an indication for highlighting the position of the object is superimposed, and (a3) outputs the display image.

Abstract: Disclosed is a method for planning screw locking path using an ant colony algorithm, which includes: obtaining designated positions of screw holes to be locked; using a distance between the screw holes to be locked as pheromone; obtaining a set of initial paths for all lockings; determining whether a condition for ending an iteration is met, and the condition is whether all the locking paths have passed through all the designated positions; if the condition for ending the iteration is not met, obtaining a supplementary path for each locking path to form an entire path of each locking path until the condition for ending the iteration is met; taking a set of entire paths of all the locking paths as a set of final paths; obtaining a shortest path from the set of final paths.

Abstract: A system and method for semantic segmentation with a soft cross-entropy loss is provided. The system inputs a first color image to an input layer of a semantic segmentation network for a multi-class classification task. The semantic segmentation network generates, at an auxiliary stride, a first feature map as an output of an auxiliary layer of the semantic segmentation network based on the input first color image. The system extracts the generated first feature map from the auxiliary layer and computes a probability map as a set of soft labels over a set of classes of the multi-class classification task, based on the extracted first feature map. The system further computes an auxiliary cross-entropy loss between the computed probability map and a ground truth probability map for the auxiliary stride and trains the semantic segmentation network for the multi-class classification task based on the computed auxiliary cross-entropy loss.

Abstract: A system and method for star tracking includes: capturing an image of stars; detecting and selecting visible stars from the captured image; extracting features from the selected stars by forming a convex hull from the selected stars to generate a spherical polygon; computing the area and higher order moments of the spherical polygon; and pattern matching the extracted feature against a database of star catalog. The pattern matching includes matching the area of the spherical polygon to a plurality of polygon areas stored in the database and when the number of the matching candidates is more than one, matching a next extracted higher order moment with a respective higher order moment in the database, and repeating said matching of the next extracted higher order moment until the number of the matching candidates is equal to one.

Abstract: First and second confidence values are determined. The first confidence value corresponds to a likelihood that, from a position and an orientation of a mobile computing device, the mobile computing device is pointing to a visual code. The second confidence value corresponds to a likelihood that, from an image captured by the mobile computing device, the mobile computing device is pointing to the visual code. Whether the mobile computing device is pointing to the visual code is determined based on the first confidence value and the second confidence value.

Abstract: A method and apparatus for detecting a moving target, an electronic equipment and a storage medium. The method includes obtaining a first frame image and a second frame image which are adjacent, and a rotation matrix and a translation matrix between the first and second frame images, the first and second frame image include the same moving target; extracting first feature points from the first frame image; determining second feature points corresponding to the first feature points from the second frame image based on the second frame image and the first feature points; determining distances between the second feature points and corresponding polar lines based on the rotation matrix and the translation matrix; determining third feature points located on the moving target based on the distances; and detecting the moving target based on the third feature points.

Abstract: A method of identifying contiguities in images is disclosed. The contiguities are indicative features and various qualities of an image, which may be used for identifying objects and/or relationships in images. Alternatively, the contiguities may be helpful in ensuring that an image has a desired switching factor, so as to create a desired effect when combined with other images in a composite image. The contiguity may be a group of picture elements that are adjacent to one another that form a continuous image element that extends generally horizontally (e.g., diagonally, horizontally) across the image.

Abstract: A method for monitoring a gauge uses a machine learning model that was previously trained based on a set of training data to classify relative indicator positions between any pair of gauge images. The method receives an input indicative of a first reference gauge value and a first gauge image. The method classifies, using the machine learning model, the indicator position on the first gauge image relative to the indicator position on a first reference gauge image associated with the first reference gauge value and provides an output having a first value indicative of the indicator position of the first gauge image being on a first side of the indicator position of the first reference gauge image and a second value indicative of the indicator position of the first gauge image being on a second side of the indicator position of the first reference gauge image.

Abstract: System and techniques for calibrating a crop row computer vision system are described herein. An image set that includes crop rows and furrows is obtained. Models of the field are searched to find a model that best fits the field. A calibration parameter is extracted from the model and communicated to a receiver.

Abstract: Embodiments provide multi-angle screen coverage analysis. In some embodiments, a system obtains at least one image, where the at least one image is a computer graphics generated image, and where the at least one image comprises at least one target object. The system determines screen coverage information for the at least one target object, where the screen coverage information is based on a portion of the screen that is covered by the at least one target object. The system determines depth information for the at least one target object. The system determines an asset detail level for the at least one target object based on the screen coverage information and the depth information. The system then stores the asset detail level in a database.

Abstract: A lane recognition device includes: a camera configured to capture an image in front of a vehicle; and a controller configured to detect a lane from the image in front of the vehicle; generate a plurality of lane equations based on a curved point of the lane; and recognize the lane based on the plurality of lane equations.

Abstract: A method implemented on a computing device comprising a data-parallel coprocessor and a memory coupled with the data-parallel processor for generating and evaluating N-to-1 mappings between spatial point sets in nD. The method comprises using the computing device to carry out steps comprising receiving a first and a second spatial point sets, an array of (n+1) combinations in the first spatial point set, an array of one or more pairs of neighbor (n+1) combinations referencing into the array of (n+1) combinations, and a CCISS between the two spatial point sets; computing a plurality of solving structures and provide a two-level indexing structure for (n+1) combinations for the plurality of (n+1) combinations; generating one or more N-to-1 mappings; and generating a plurality of local distance measures for unique combinations of the one or more pairs of neighbor (n+1) combinations and the one or more N-to-1 mappings.

Abstract: According to an embodiment, a captured image check system includes a camera, an image data input unit, a camera view angle judge processor, a road surface state analysis judge processor, and an output unit. The camera photographs a road surface. The image data input unit inputs image data to be used for analyzing a state of the road surface. The camera view angle judge processor performs a judge process of determining whether or not an angle of view of the camera satisfies a first condition for analyzing the state of the road surface on the basis of the image data. The road surface state analysis judge processor performs a judge process of determining whether or not a second condition for analyzing the state of the road surface is satisfied on the basis of image data of the road surface. The output unit outputs results of the judge processes.

Abstract: A method for detecting dirt on a screen, a device for the same, an electronic device and a computer-readable storage medium are provided. The method includes acquiring a first image of the screen when the screen is displaying a first picture, determining a dirt detection area in the first image with a mask image, and detecting a position of the dirt within the dirt detection area. With the method provided by the embodiments of the disclosure, the positon of dirt on the screen could be detected based on at most two images of the screen. The time for detecting dirt on the screen is greatly shortened, and the efficiency of detection is enhanced.

Abstract: A method of identifying and analyzing contiguities in images is disclosed. The contiguities are indicative features and various qualities of an image, which may be used for identifying objects and/or relationships in images. Alternatively, the contiguities may be helpful in ensuring that an image has a desired figure-ground ambiguous switch between percepts, so as to create a desired effect when combined with other images to generate a composite image set. The contiguity may be a group of picture elements that are adjacent to one another that form a continuous image element that extends generally horizontally (e.g., diagonally, horizontally and/or vertically) across the image.

Abstract: Methods and systems are provided for profile-based selection of imaging parameters for a clinical context. In one embodiment, a method comprises selecting a profile based on an indication of clinical context for a scan of a patient, selecting one or more sub-profiles based on the profile, selecting imaging settings for the scan based on the profile and the one or more sub-profiles, and performing the scan according to the selected imaging settings. In this way, a scan prescription may be defined more efficiently and personally for a patient while improving quality assurance and managing an x-ray radiation dose of the patient.

Abstract: A method of verifying a fingerprint includes selecting a feature point from among feature points of a fingerprint image, the selected feature point matching a registered feature point; determining a partial region of the fingerprint corresponding to the feature point in the fingerprint image; obtaining feature information of the partial region of the fingerprint in the fingerprint image; and verifying the fingerprint based on the feature information of the partial region in the fingerprint image.

Abstract: The present invention relates to a method for detecting a marker and an electronic device thereof. According to various embodiments of the present invention, the electronic device comprises: at least one sensor; a display; a camera; at least one processor connected with the at least one sensor, the display, and the camera; and a communication circuit coupled with the at least one processor. The at least one processor acquires a first image including a marker displayed on another electronic device through the camera, determines a form of the marker on the basis of analysis information of the first image, and can be set so as to transmit information on the form of the marker to the other electronic device through the communication circuit. Other embodiments are also possible.

Abstract: Combinatorial shape regression is described as a technique for face alignment and facial landmark detection in images. As described stages of regression may be built for multiple ferns for a facial landmark detection system. In one example a regression is performed on a training set of images using face shapes, using facial component groups, and using individual face point pairs to learn shape increments for each respective image in the set of images. A fern is built based on this regression. Additional regressions are performed for building additional ferns. The ferns are then combined to build the facial landmark detection system.

Abstract: A method of determining dynamic properties of a structure (linear or nonlinear) includes receiving spatio-temporal inputs, generating mode shapes and modal components corresponding to the spatio-temporal inputs using a trained deep complexity coding artificial neural network, and subsequently generating the dynamic properties by analyzing each modal component using a trained learning machine. A computing system for non-contact determination of dynamic properties of a structure includes a camera, a processor, and a memory including computer-executable instructions. When the instructions are executed, the system is caused to receive spatio-temporal image data, decompose the spatio-temporal image data into constituent manifold components using an autoencoder, and analyze the constituent manifold components using a trained learning machine to determine the dynamic properties.