Abstract: Systems and methods for identifying transformations to be applied to at least part of a document image for improving the OCR quality. An example method comprises: constructing, by a computer system, an ordered list of transformations to be applied to an image comprising a character string, each transformation corresponding to a hypothesis asserted with respect to one or more characteristics of the image; applying, to the image, a leading transformation on the list to produce a transformed image; evaluating a quality of the transformed image to produce a quality estimate; and updating the list in view of the quality estimate.
Abstract: A contextual local image recognition module of a device retrieves a primary content dataset from a server and then generates and updates a contextual content dataset based on an image captured with the device. The device stores the primary content dataset and the contextual content dataset. The primary content dataset comprises a first set of images and corresponding virtual object models. The contextual content dataset comprises a second set of images and corresponding virtual object models retrieved from the server.
Abstract: Techniques for enhancing an image using pixel-specific processing are disclosed. An image can be enhanced by updating certain pixels through patch aggregation. Neighboring pixels of a selected pixel are identified. Respective patch values for patches containing the selected pixel are determined. Patch values provide update information for updating the respective pixels in the patch. Relevant patch values for the selected pixel are identified by identifying associated patches of the pixel. Information from the relevant patch values of the selected pixel may be obtained. Using this information, pixel-specific processing may be performed to determine an updated pixel value for the selected pixel or for neighboring pixels of the selected pixel. Pixel-specific processes may be executed for each of the selected or neighboring pixels. These pixel-specific processes can be executed in parallel.
Abstract: In a case that a pixel shape in a luma domain is not similar to a pixel shape in a chroma domain in an input image, a prediction image that is generated in the reuse of a luma prediction mode is made more accurate. A moving image decoding apparatus 1 includes a gradient deriving unit 1453B that references a chroma gradient definition DEFANG1C, and derives a chroma prediction direction from a prediction mode based on a luma prediction direction.
Abstract: When there is a possibility that a third character region is redundantly selected in both a case where the line extraction process is performed starting from a first character region and a case where the line extraction process is performed starting from a second character region located in a line different from a line containing the first character region, the line recognition unit determines which line to incorporate the third character region in, by comparing a case of incorporating the third character region into the line starting with the first character region, with a case of incorporating the third character region into the line starting with the second character region.
Abstract: An image processing apparatus that is capable of improving image stabilization performance by improving motion vector search performance by restricting a search range even for an image with a large motion due to a camera shake of an image pickup apparatus at a tele-end side of a zoom lens or a shoot while walking at a wide-end side. A motion vector search unit searches for a motion vector of a reference image later input from an image pickup device with respect to a standard image input previously. A control unit sets at least one of a search range and a search position of a motion vector searched by the motion vector search unit based on position-posture change information about the image pickup apparatus detected by a detection unit and image stabilization control information obtained from an image stabilizing unit.
Abstract: A feature extraction method includes acquiring an image, detecting a vanishing point in the image, setting a plurality of ellipses and a plurality of half-lines with respect to the detected vanishing point to segment the image into a plurality of regions, and extracting a feature of each segmented region. With the disclosed method, a global feature can be appropriately extracted from an image including a vanishing point or other relevant features.
Abstract: According to an embodiment of the present invention, an imaging apparatus includes an image acquiring unit, an image correcting unit, and a correction preprocessing unit. The image acquiring unit acquires an image. The image correcting unit performs distortion correction for the image acquired by the image acquiring unit. The correction preprocessing unit determines an inclination of a face of a person in the image acquired by the image acquiring unit. Further, the correction preprocessing unit performs control so that the image correcting unit performs distortion correction of the image at an intensity according to the determined inclination of the face.
Abstract: An image processing device includes a gatherer configured to receive image data and gather statistics of color data from the image data. The image processing device includes a processor configured to analyze the statistics of the color data to determine whether the image data includes single color data or multi color data, determine a first gain value to perform an automatic white balance if the image data includes multi color data, and determine a second gain value different from the first gain value using a gain table if the image data includes single color data. The image processing device includes a gain applier configured to apply the first gain value or the second gain value to the color data.
Abstract: A transmission device divides image data into a plurality of regions on the basis of a similarity of pixels and transmits the image data and region representative points of the plurality of regions. A reception device receives the image data and the region representative points transmitted from the transmission device. The reception device generates region labels used for identifying the plurality of regions using the image data and the region representative points.
Abstract: The image coding method includes: determining a context for a current block to be processed, from among a plurality of contexts; and performing arithmetic coding on the control parameter for the current block to generate a bitstream corresponding to the current block, wherein in the determining: the context is determined under a condition that control parameters of neighboring blocks of the current block are used, when the signal type is a first type, the neighboring blocks being a left block and an upper block of the current block; and the context is determined under a condition that the control parameter of the upper block is not used, when the signal type is a second type, and the second type is one of “mvd_l0” and “mvd_l1”.
Abstract: A selected color component (R, G, B) for each of a plurality of image pixels (32, P) of an image cell (42, C) is encoded to form a corresponding color component (ER, EG, EB) of a corresponding plurality of encoded pixels (40, PE) in one-to-one correspondence with the image pixels (32, P). A bit length of each color component (ER, EG, EB) of an encoded pixel (40, PE) is less than that of a corresponding image pixel (32, P).
Abstract: In various embodiments, a system and method for manufacturing and using an optical enhancement assembly in combination with an image-capturing device are presented. In example embodiments, the optical enhancement assembly is affixed over an aperture on the image-capturing device using a securing agent. Light is allowed to travel through the optical enhancement assembly and into the aperture of the image-capturing device such that the light can be recorded as a still image or video. The optical enhancement assembly includes at least a unique fractalized diffraction pattern that impacts light traveling into the aperture and causes a unique diffraction effect on the image or video recorded by the image-capturing device.
Abstract: Feature extraction of image data using feature extraction modules. The feature extraction modules may be provided in an architecture that allows for modular, decoupled generation and/or operation of the feature extraction modules to generate feature data corresponding to image data. In this regard, the feature extraction modules may communicate with a file system storing image data and feature data by way of a common interface format. Accordingly, regardless of the nature of the execution of the feature extraction module, each feature extraction module may be communicative by way of the common interface format, thereby providing a modular approach that is highly scalable, flexible, and adaptive.
Type:
Grant
Filed:
August 14, 2015
Date of Patent:
February 14, 2017
Assignee:
DigitalGlobe, Inc.
Inventors:
Giovanni B. Marchisio, Mark D. Tabb, Carsten Tusk, Krzysztof Koperski, Jeffrey D. Shafer
Abstract: Described herein is a technology for facilitating remote monitoring, in accordance with one aspect, image data and corresponding true color data of a region of interest is received by a computer system from a mobile device. The computer system may integrate the image data and the true color data to generate normalized true color data. The normalized true color data may then be mapped to device independent color image data. A recommendation may then be sent based on the device-independent color image data.
Abstract: Medical thermal image processing for vein or other subsurface element detection and related methods is provided. A method for thermal image processing to enhance thermal scene information includes taking an output of an infrared camera and coupling the output to a spatial high pass filter. An output of the spatial high pass filter is then applied to a scene-based noise suppression module.
Type:
Grant
Filed:
October 16, 2014
Date of Patent:
January 24, 2017
Assignee:
BAE Systems Information and Electronic Systems Integration Inc.
Abstract: To improve an image recognition rate by quickly changing a parameter in a proper manner without being affected by a transmission delay of an image encoding stream in an image receiving device that recognizes a decoded image obtained by decoding the received image encoding stream. The image receiving device includes a data receiving unit, a parameter changing unit, a decoding unit, and an image recognition unit. The data receiving unit receives an image encoding stream including image encoding data and the parameter. The parameter changing unit changes the parameter received by the data receiving unit, that is, the parameter specified for encoding performed by a sender, to a value suitable for image recognition performed in the subsequent stage. The decoding unit generates the image decoding data by decoding the received image encoding data according to the changed parameter. The image recognition unit performs image recognition on the image decoding data.
Abstract: A method for proactively creating an image product includes storing a library of specification terms for image products by a computer system, receiving a command from a user, tokenizing the command into a plurality of tokens by the computer system, matching one or more of the tokens to the specification terms in the library to determine specification parameters for an image product by the computer system, automatically identifying images based on the specification parameters by the computer system, and automatically creating a design for the image product that incorporates at least some of the images identified based on the specification parameters.
Type:
Grant
Filed:
January 7, 2016
Date of Patent:
January 17, 2017
Assignee:
Shutterfly, Inc.
Inventors:
Eugene Chen, Preeti Nathan, Trynne Anne Miller, Wiley H. Wang, Shay Litvak, Marco Santini, Chris M. Denend
Abstract: An image processing device for correcting at least a portion of a face image containing a person's face, includes an image acquisition unit that acquires a face image, a region extractor that extracts a pupil region containing a pupil from the face image, a determination unit that determines a presence or absence of halation in the pupil region, a first template acquisition unit that acquires a first template image which is a correction template image, a reference color acquisition unit that acquires a reference color corresponding to the person from the face image, a second template acquisition unit that generates a second template image using the first template image and the reference color, and a correction unit that composites the second template image with at least the region in the pupil region where the halation occurred to correct the halation.
Abstract: A method for determining user preferences for data visualizations is provided. The method may include receiving data visualizations. The method may also include collecting the shapes, the line segments, and the colors associated with the data visualizations. The method may further include converting the shapes and the line segments to polygonal outlines. Additionally, the method may include categorizing and measuring the line segments. The method may further include identifying and categorizing the angles formed by the line segments and determining weighted values for the angles. The method may further include calculating the total length for the line segments, and the total weighted value for the angles. The method may also include characterizing the line segments based on the categorization of the line segments and the angles based on the categorization of the angles. The method may further include scoring the at least one data visualization based on the characterizations.
Type:
Grant
Filed:
June 29, 2015
Date of Patent:
January 10, 2017
Assignee:
International Business Machines Corporation
Inventors:
T. Alan Keahey, Daniel J. Rope, Graham J. Wills