Abstract: A method of detecting a blemish in a digital camera includes capturing a digital image of a white area by the digital camera; reducing resolution of the digital image by grouping pixels of the digital data into a matrix of a plurality of data blocks; and comparing a luminance value of each of the plurality of data blocks to an average luminance value of other data blocks of the plurality of data blocks within a predetermined region to determine if each of the plurality of data blocks is a potentially bad data block.

Abstract: An inspection machine includes image sensors that image a sheet, a lighting device that irradiates the sheet with light, and an image processing device that generates an image of an object from outputs of the image sensors. The image processing device generates captured images that include an image in a predetermined range of the sheet in common based on the outputs from the image sensors. The image processing device extracts images of the object on the sheet for each of the generated captured images. The image processing device associates the extracted images of the object with each other based on a feature amount of the object reflected in the images. The image processing device combines the associated images of the object with each other.

Abstract: The following relates generally to light detection and ranging (LIDAR) and artificial intelligence (AI). In some embodiments, a system: receives LIDAR data generated from a LIDAR camera; determines data of a first object based upon processor analysis of the LIDAR data, the data of the first object comprising: (i) dimensional data of the first object, and (ii) a type of the first object; and adds the first object and the first object data to a commercial inventory list.

Abstract: The invention provides an English char image recognition method mainly generating a rectangular coordinate frame from a loaded English character image, finding a gravity center in the rectangular coordinate frame, obtaining feature points on the rectangular coordinate frame and around the gravity center, performing one-dimensional convolutional operation and processing on the feature points to generate six layer feature maps, and generating a character recognition result from the six layer feature maps to solve the problem of generating a large amount of computation in conventional two-dimensional recognition operation, thereby achieving an efficacy of reducing recognition equipment costs and enabling fast and accurate recognition.

Abstract: A recording medium determination apparatus includes an image data acquisition unit configured to acquire image data obtained by capturing an image of a predetermined area in a recording medium, a first extraction unit configured to extract a first feature amount by processing the image data using a first parameter, a second extraction unit configured to extract a second feature amount by processing the image data using a second parameter different from the first parameter, and a determination unit configured to determine a type of the recording medium based on the first feature amount and the second feature amount.

Abstract: Disclosed herein are system, method, and computer program product embodiments for detecting a potential damage caused to a vehicle when the vehicle is placed on a transporter. The method includes obtaining a first image and a second image taken at different time instances. The first image and the second image are generated by one or more cameras placed around a vehicle on a transporter, and cover a same portion of a body of the vehicle. The method further includes detecting a change between a portion of the first image corresponding to the portion of the body of the vehicle and a portion of the second image corresponding to the portion of the body of the vehicle indicative of a collision. Moreover, the method includes flagging a potential damage to the portion of the body of the vehicle based on the detected change between the first image and the second image.

Abstract: A medical image processing apparatus includes processing circuitry configured to acquire medical image data representing a region including a target site, divide the medical image data into a plurality of pieces of divisional data according to an anatomical structure, extract a target region corresponding to the target site from each of the pieces of divisional data according to a particular condition, and cause an output circuit to output information on the extracted target region.

Abstract: The present invention proposes a technique for enabling the execution of measurement processing without referring to a design drawing for which it is difficult to adjust or obtain parameters for image processing that requires knowhow. This measurement system according to the present disclosure refers to a learning model generated on the basis of teaching data, which is generated from a sample image of a semiconductor, and the sample image, generates a region-segmented image from an input image (measurement subject) of a semiconductor having a predetermined structure, and uses the region-segmented image to perform image measurement. Here, the teaching data is an image in which labels, which include a structure of the semiconductor in the sample image, are assigned to each pixel of the image, and the learning model includes parameters for deducing teaching data from the sample image (see indicator 1).

Abstract: A method for determining images plausible to have a false negative object detection comprises providing a group of historic trajectories, wherein each historic trajectory comprises a reference track that represents one or more historic tracks and comprises an object class of historic object detections that belong to the one or more historic tracks; performing tracking; performing object detection; for a determined track that does not match any determined object detection, comparing the determined track with reference tracks of historic trajectories for identifying a matching reference track; upon identifying a matching reference track, defining images of the determined track as being plausible to have a false negative object detection for the object class of the historic trajectory comprising the matching reference track; and upon not identifying a matching reference track, defining the determined track as a false positive track.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed to analyze characteristics of text of interest using a computing system. An example apparatus includes a text detector to provide text data from a first image, the first image including a first text region of interest and a second text region not of interest, a color-coding generator to generate a plurality of color-coded text-map images, the plurality of color-coded text-map images including color-coded segments with different colors, the color-coded segments corresponding to different text characteristics, and a convolutional neural network (CNN) to determine a first location in the first image as more likely to be the first text region of interest than a second location in the first image corresponding to the second text region that is not of interest based on performing a CNN analysis on the first image and the plurality of color-coded text-map images.

Abstract: A vehicle occupant gaze detection system includes a non-transitory computer readable medium configured to store instructions thereon; and a processor connected to the non-transitory computer readable medium. The processor is configured to execute the instructions for receiving gaze data from a vehicle. The gaze data includes a viewing direction of an occupant of the vehicle, vehicle information, and time stamp information. The processor is configured to execute the instructions for generating a gridmap including an array of lattice points, and each lattice point of the array of lattice points corresponds to a location. The processor is configured to execute the instructions for generating a histogram including information related to viewing of at least one location by the occupant. The processor is configured to execute the instructions for determining an effectiveness of an object based on the histogram; and transmitting a recommendation based on the effectiveness of the object.

Abstract: The embodiments of the present disclosure disclose image defogging method. A specific mode of carrying out the method includes the steps to: obtain a polarized fog-free scene image set, generate a first quintuple based on each polarized fog-free scene image, and obtain a first quintuple set; extract pixel coordinates meeting a preset condition from the significant polarization image included in each first quintuple, to obtain a pixel coordinate group to generate a second quintuple; generate a simulated polarization foggy scene image sequence based on each second quintuple, to obtain a simulated polarization foggy scene data set; design a polarization state attention neural network; obtain a polarized foggy scene image, and input the polarized foggy scene image into the polarization state attention neural network to obtain a polarized defogging image. This embodiment can still improve the accuracy of image defogging with poor lighting conditions.

Abstract: Methods, systems, and apparatus for high-fidelity vision tasks using deep neural networks are disclosed. An example apparatus includes a feature extractor to extract low-level features and edge-enhanced features of an input image processed using a convolutional neural network, an eidetic memory block generator to generate an eidetic memory block using the extracted low-level features or the extracted edge-enhanced features, and an interactive segmentation network to perform image segmentation using the eidetic memory block, the eidetic memory block used to propagate domain-persistent features through the segmentation network.

Abstract: Provided are template mark detection method and template position correction method based on single camera, including: performing image collection on mark on template by single camera, and obtaining binary image after preprocessing; performing corner detection of jagged edges on binary image to obtain corner set of jagged edges; performing edge detection and line detection sequentially on binary image to obtain set of edge line segments from coarse detection; traversing such set, and judging and retaining collinear line segments, to obtain set of collinear line segments from coarse detection; traversing corner set of jagged edges, for point-line collinearity judgment with line segments in set of collinear line segments from coarse detection, to obtain set of point-line from fine detection; and performing linear fitting on set of point-line from fine detection, and calculating an inclination angle of each straight line through an arctan function, thus completing detection of mark.

Abstract: A method including generating a captured facial object and a captured pose from a captured image. The method also includes obtaining a base facial object and a base pose from a base image. The method also includes generating base pose angles using the captured pose, and captured pose angles using the captured pose. The method also includes obtaining selected base images using the base pose angles and the base facial object. The method also includes generating selected captured images using the captured pose angles and the captured facial object. The method also includes comparing the selected base images to the selected captured images to establish a comparison. The method also includes outputting a match output using the comparison.

Abstract: In one embodiment, a method includes accessing a stream of F video frames, where each of the F video frames includes N patches that are non-overlapping, generating an initial embedding vector for each of the N×F patches in the F video frames, generating a classification embedding by processing the generated N×F initial embedding vectors using a self-attention-based machine-learning model that computes a temporal attention and a spatial attention for each of the N×F patches, and determining a class of the stream of video frames based on the generated classification embedding.

Abstract: The invention relates to a process for identification of an implant, optionally worn by an individual, comprising the following steps, performed by a processing unit: obtaining a radiographic input image in which a dental implant is visible; classification of a region of interest of said radiographic image by means of at least one convolutional neural network; said classification producing a list of candidate implants to be the implant visible on the image, the list being ordered as a function of a probability of being the implant visible on the image.

Abstract: Human lying posture detections are disclosed. A human lying on a bed is detected by obtaining an image through a depth camera, detecting objects in the image and marking the objects in the image using 2D bounding boxes by deep learning, determining the human being in a lying posture in response to a width and a height of the 2D bounding box of the human meeting a predetermined condition, detecting one or more skin areas in the image and generating skin area 2D bounding boxes to mark each of the one or more skin areas using a skin detection algorithm, and determining the human being in the lying posture in response to the skin area 2D bounding boxes and the 2D bounding box of the bed meeting a predetermined positional relationship.

Abstract: Various methods and systems are provided for x-ray imaging. In one embodiment, a method includes acquiring, with an x-ray detector, an x-ray image of a subject, determining a transformation that minimizes anti-scatter grid artifacts in the x-ray image, correcting the x-ray image according to the transformation to generate a corrected image, and outputting the corrected image. In this way, artifacts arising from a misalignment of an anti-scatter grid between the calibration and the acquisition may be reduced.

Abstract: By using a plurality of images obtained by combining image of damage included in a road surface still image based on perspective and definitions of the damage as damage learning data, an object detection model is built. A plurality of data wherein the number of rectangles in an analysis result image displaying the rectangles surrounding the damage output as an analysis result of the object detection model and a pavement condition rating result obtained by visual inspection of the analysis result image of the object detection model are combined is used as condition rate learning data to build a rating estimation model, and pavement condition is rated using the rating estimation model.