The invention provides an ultrasound data processing method (30) for detecting presence of an intravascular object in a vessel lumen based on analysis of acquired intravascular ultrasound data of the lumen. The method comprises receiving (32) data comprising multiple frames, and each frame containing data for a plurality of radial lines, corresponding to different circumferential positions around the IVUS device body, and reducing (34) the data to a single representative value for each radial line in each frame. These representative values are subsequently processed to derive (36) values for at least each frame representative of a probability of presence of an object within the given frame. Based on the probability values, a region within the data occupied by an intravascular object, for instance a consecutive set of frames occupied by an object, is determined (38).

Methods, systems, computer-readable storage media, and apparatuses for efficiently computing a computer vision operation are presented. In certain aspects, techniques are disclosed for receiving values from one or more pixels from a pixel array and representing those values for facilitating computer vision operations, such as Haar-like feature computations. In one implementation, the pixel values are represented as a hierarchy of computations; each level above the lower level of the hierarchy of computations comprises one or more values representing a computation of the values from a lower level of the hierarchy. The computation can include a simple sum of the values for the lower level of the hierarchy, a weighted sum of the values for the lower level of the hierarchy, an average of the values for the lower level of the hierarchy, a local binary pattern based on the values for the lower level of the hierarchy, a histogram of oriented gradients based on the values for the lower level of the hierarchy, or other computer vision feature computation based on the values for the lower level of the hierarchy. In such an implementation, the computer vision operations, such as Haar-like feature computations, may be performed using the hierarchy of averages.

In a video conference system in which at least a pair of terminal devices transmits and receives an image through a network, each of the terminal devices includes a face detection unit that detects a face from a first image which is a image captured by a camera, and a generation unit that generates a image in which a image of the face detected by the face detection unit is arranged around a image region of a second image, which is a image of a material used for a conference, in accordance with a relative position of the face in the first image.

A method of tracking faces in an image stream with a digital image acquisition device includes receiving images from an image stream including faces, calculating corresponding integral images, and applying different subsets of face detection rectangles to the integral images to provide sets of candidate regions. The different subsets include candidate face regions of different sizes and/or locations within the images. The different candidate face regions from different images of the image stream are each tracked.

Disclosed is an image feature extraction method including a step of defining a combination of at least two kinds of solid angles along at least two directions, of an input spherical image; a step of determining respective values of the combination of the at least two kinds of solid angles, so that surface areas of spherical crowns of spherical segments, which are obtained by dividing the spherical image by the respective values of the combination of the at least two kinds solid angles, have a same value; and a step of generating, by utilizing the respective values of the combination of the at least two kinds of solid angles, an image feature template so as to conduct image feature extraction.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for facial recognition using fractal features are disclosed. In one aspect, a method includes the actions of accessing data encoding a facial image, the facial image including a face. The actions further include generating a hierarchical graphical model of the face in the facial image, the hierarchical graphical model including more than nodes from at least one level, each level approximating the face within a contour. The actions further include applying a bank of filters to the face at a particular node, each filter spanning more than one scale and at least one direction. The actions further include analyzing filter responses from the bank of filters applied at each direction to obtain a similarity measure that consolidates filter responses from filters applied at more than one scale. The actions further include generating a vector representation.

A method and system for approximating a deep neural network for anatomical object detection is discloses. A deep neural network is trained to detect an anatomical object in medical images. An approximation of the trained deep neural network is calculated that reduces the computational complexity of the trained deep neural network. The anatomical object is detected in an input medical image of a patient using the approximation of the trained deep neural network.

The present invention relates to a user authentication method, a device for executing the same, and a recording medium for storing the same. A user authentication method executed in a user authentication device according to an embodiment of the present invention comprises: a step of, when image data of a user is received from an image photographing device, detecting a facial area and a facial feature point using each frame image of the image data; a step of performing a face authentication by matching the facial area with a predetermined face template; a password authentication step of detecting an eye winking using an image of an eye area extracted using the facial feature point, recognizing a password according to a state of the eye winking on the basis of a preconfigured reference, and checking whether the recognized password matches with a preconfigured password; and a step of determining that the user authentication is successful, on the basis of the results obtained from the face authentication and the password authentication.

A digital camera has an integral flash and stores and displays a digital image. Under certain conditions, a flash photograph taken with the camera may result in a red-eye phenomenon due to a reflection within an eye of a subject of the photograph. A digital apparatus has a red-eye filter which analyzes at least one partial face region identified within the digital image for the red-eye phenomenon and modifies the image to eliminate the red-eye phenomenon by changing the red area to black. The modification of the image is enabled when a photograph is taken under conditions indicative of the red-eye phenomenon. The modification is subject to anti-falsing analysis which further examines the area around the red-eye area for indicia of the eye of the subject. The detection and correction can be optimized for performance and quality by operating on subsample versions of the image when appropriate.

Subject matter disclosed herein relates to arrangements and techniques that provide for identifying objects within an image such as the face position of a user of a portable electronic device. An application specific integrated circuit (ASIC) is configured to locate objects within images. The ASIC includes an image node configured to process an image and a search node configured to search the image for an object in the image. The search node includes an integral image generation unit configured to generate an integral image of the image and a Haar feature evaluation unit configured to evaluate search windows of the integral image with respect to Haar-like features. The ASIC also includes an ensemble node configured to confirm the presence of the object in the image.