Abstract: In the present disclosure, a candidate area is determined based on a pixel having a specific color included in an input image, and an area is determined to be a processing target from the candidate area based on a pixel having a predetermined color different from the specific color included in the candidate area. Further, a second binary image in which a pixel corresponding to the pixel having the specific color is converted into a white pixel is generated by converting, in a first binary image obtained by the input image being binarized, a pixel that is included in the area determined to be the processing target and corresponds to the pixel having the specific color, into a white pixel.

Abstract: An object included in a low-resolution image can be recognized with a high degree of precision. An acquisition unit acquires, from a query image, an increased-resolution image, which is acquired by increasing the resolution of the query image, by performing pre-learned acquisition processing for increasing the resolution of an image. A feature extraction unit, using the increased-resolution image as input, extracts a feature vector of the increased-resolution image by performing pre-learned extraction processing for extracting a feature vector of an image. A recognition unit recognizes an object captured on the increased-resolution image on the basis of the feature vector of the increased-resolution image and outputs the recognized object as the object captured on the query image.

Abstract: Provided are computer-implemented technologies of authenticating documents. The technologies use a set of photo images (or videos), taken under a certain illumination condition and from a set of distinct tilting angles, on one or more security/ID features of one or more documents of a document genre whose authenticity is ascertained, to train an Artificial Intelligence (AI) machine learning program to build a learned model. The learned model, through the said training, attains a set of angular responses of the document genre under the illumination condition which encode a set of descriptive information about each of the one or more security/ID features of the document genre under the illumination condition. The learned model, then, is applied to authenticate, one by one, a number of target documents of the document genre whose authenticity is unknown and to be determined.

Abstract: A liveness detection system uses a combination of interactive and silent tests to verify physical presence of a live person at a client device. The system superimposes a box on a live video of the space in front of the client device and instructs the user to move to a position that places his or her face within the box. While the user's face is within the box, the system captures a sequence of frames from the video while an amount of illumination cast on the user's face is switched between bright and dark levels according to a random, semi-random, or predefined flash sequence. The system analyzes the resulting sequence of frames to confirm that the user's head pose is consistent with the location of the box, and that changes in brightness of the image of the user's face across the captured frames are consistent with the flash sequence.

Abstract: An object detection system and method includes: an optical image sensor arranged to perform the following steps: capturing a calibration image during a calibration stage, dividing the calibration image into a plurality of quadrants, and calculating a parameter for each of the quadrants; capturing a plurality of raw images during a detection stage, dividing each image of the raw images into a plurality of quadrants, and calculating a parameter for each of the quadrants; comparing the respective parameters of each quadrant of a raw image with the respective parameters of each quadrant of the calibration image to generate a ratio value for each quadrant; and comparing the ratio value of each quadrant with a predetermined threshold. When each ratio value of specific quadrants of the quadrants is greater than the predetermined threshold, object detection is confirmed.

Abstract: What is disclosed are systems and methods for optical correction for correcting for non-uniformity in active matrix light emitting diode device (AMOLED) and other emissive displays, using iterative processing of images of calibration patterns including features of coarse and fine granularity to successively generate a high-resolution estimate of the panel pixel locations.

Abstract: A driver monitor system and method of optimizing use of cognitive resources of a user in a vehicle is disclosed. The system includes in-cabin video cameras, a display device, and processing circuitry. The method performed by the processing circuitry includes monitoring eye gaze direction using images from the in-cabin video cameras, gradually softening image regions and lines displayed in the display device that are outside a first vision area of the user of the vehicle, and using the eye gaze direction to predict whether the user is transitioning into a new focus vision area that had been outside the first vision area. When the user's eyes begin to move, begin sharpening softened image regions and lines such that by the time eyes have shifted into the new focus vision area, the image regions and lines in the new focus vision area reach full sharpness.

Abstract: Disclosed herein are systems and methods for analyzing one or more package. In an embodiment, disclosed is a method comprising determining, by a sensor component of a package analysis device comprising a processor that executes computer executable components stored in a memory and the memory that stores computer executable components, a presence of a package and a set of dimensions of the package and extracting, by an image capturing component of the package analysis device, a set of image data from a package label based on a determined presence of the package.

Abstract: Stubble burning is a serious problem resulting in pollution attributable to smog, loss of nutrients in the top soil, and risk of fires spreading out of control. Existing methodologies have attempted to predict burning areas, but have failed to do so because of inefficient sensing mechanism. Present disclosure proposes a system and method to compute burning index score pertaining to crops by detecting harvest season and predicting probable areas of burning by combining current year's crop area map along with harvesting period and historical hot spot information. Computation of the burning index score is accomplished based on inputs received from at least one of satellite imaging, multi-spectral drone based sensing devices and crowdsourcing information. This will help to prioritize the area for taking corrective measures such as training of farmers, availability of resources, optimizing the resources schedule, etc.

Abstract: Systems and methods are disclosed for receiving a target image corresponding to a target specimen, the target specimen comprising a tissue sample of a patient, applying a machine learning model to the target image to determine at least one characteristic of the target specimen and/or at least one characteristic of the target image, the machine learning model having been generated by processing a plurality of training images to predict at least one characteristic, the training images comprising images of human tissue and/or images that are algorithmically generated, and outputting the at least one characteristic of the target specimen and/or the at least one characteristic of the target image.

Abstract: The present disclosure provides a method and apparatus of estimating a road condition, and a method and apparatus of establishing a road condition estimation model, which relates to a field of big data and intelligent traffic. The method includes: acquiring, for a first preset duration before a first moment, a sequence of user tracks for a road and a sequence of road images for the road; extracting a track-related feature of the road from the sequence of the user tracks, and extracting an image-related feature of the road from the sequence of the road images; and inputting the track-related feature of the road and the image-related feature of the road into a pre-trained road condition estimation model, so as to determine, for a second preset duration after the first moment, a road condition information of the road by using an estimated result of the road condition estimation model.

Abstract: Embodiments provide image display systems and methods for one or more camera calibration using a two-sided diffractive optical element (DOE). More specifically, embodiments are directed to determining intrinsic parameters of one or more cameras using a single image obtained using a two-sided DOE. The two-sided DOE has a first pattern on a first surface and a second pattern on a second surface. Each of the first and second patterns may be formed by repeating sub-patterns that are lined when tiled on each surface. The patterns on the two-sided DOE are formed such that the brightness of the central intensity peak on the image of the image pattern formed by the DOE is reduced to a predetermined amount.

Abstract: A method of predicting therapeutic effects of a treatment for cognitive impairment for an individual using the individual's predicted age difference (PAD), the method comprising scanning the individual's brain with a scanning device so as to acquire at least one medical brain image at the beginning of the treatment; processing the medical brain image to obtain at least one feature of the image; generating a PAD value of the individual based on the at least one feature of the image; comparing the PAD value with a reference value; and predicting the therapeutic effects of the treatment for cognitive impairment using the comparison result of the PAD value and the reference value.

Abstract: A method and apparatus with feature embedding is provided. The method includes estimating a depth map for each of plural two-dimensional (2D) input images, transforming the depth maps into three-dimensional (3D) information in a point cloud form based on an aggregation of the depth maps, and generating an embedded feature by applying the 3D information to a machine learning model, where the embedded feature includes information about a 3D shape corresponding to a 2D object in the plural 2D input images.

Abstract: A golf ball tracking system identifies a first pixel at which a golf ball is located in a first image frame and a second pixel at which the golf ball is located in a second image frame. The system may determine that the first pixel represents a location where the golf ball impacts the golf surface based at least in part on a difference in location between the first pixel in the first image frame and the second pixel in the second image frame.

Abstract: The present disclosure provides a method for synthesizing a projection image that represents a subject or object irradiated by X-rays from a radiation source. The method may include dividing an energy spectrum of the X-rays into one or more energy bins; determining a projection value of at least one pixel; determining a weighting coefficient corresponding to an energy bin based on a variation of the energy spectrum of the X-rays; and determining a weighted projection value of the at least one pixel based on the projection value of the at least one pixel and the weighting coefficient corresponding to the energy bin. The method may further include determining a pixel value of the at least one pixel based on the weighted projection values of the at least one pixel that correspond to all the one or more energy bins.

Abstract: A vehicular trailer assist system includes a camera disposed at a rear portion of a vehicle and having an exterior field of view that includes at least a portion of a trailer hitched to the vehicle. A control, responsive to processing of frames of image data captured by the camera and during a calibration maneuver by the vehicle, determines a trailer template of trailer hitched to the vehicle. The control, during a turning portion of the calibration maneuver, determines a trailer collision angle based on the determined trailer template. The control, after completion of the calibration maneuver, determines a current trailer angle of the trailer relative to the vehicle as the vehicle is driven along a road. The control alerts an operator of the vehicle if the current trailer angle is within a threshold amount of the determined trailer collision angle.

Abstract: A method and an apparatus for calibrating an external parameter of a camera are provided.

Abstract: Methods and systems for analyzing images are disclosed. An example method may comprise inputting one or more of a first image or a second image into a fully convolutional network, and determining an updated fully convolutional network by optimizing a similarity metric associated with spatially transforming the first image to match the second image. The one or more values of the fully convolutional network may be adjusted to optimize the similarity metric. The method may comprise registering one or more of the first image or the second image based on the updated fully convolutional network.

Abstract: Embodiments relate to a companion animal identification method including acquiring a preview image for capturing a face of a target companion animal, checking if the face of the target companion animal is aligned according to a preset criterion, capturing the face of the target companion animal when it is determined that the face of the target companion animal is aligned, and identifying the target companion animal by extracting features from a face image of the target companion animal having an aligned face view, and an identification system for performing the same.