Shape, Icon, Or Feature-based Compression Patents (Class 382/243)
-
Patent number: 11922549Abstract: Text is generated from an object. Text is generated from a first object. The first object includes a second object and a third object. A step of detecting coordinate data of the second object is included. A step of detecting coordinate data of the third object is included. A step of extracting positional relation between the second object and the third object from coordinate data is included. A step of converting the extracted positional relation into graph data is included. A step of generating text about the positional relation between the second object and the third object from graph data is included.Type: GrantFiled: July 9, 2020Date of Patent: March 5, 2024Assignee: Semiconductor Energy Laboratory Co., Ltd.Inventors: Kengo Akimoto, Junpei Momo, Takahiro Fukutome
-
Patent number: 11861884Abstract: Certain aspects of the disclosure provide systems and methods for training an information extraction transformer model architecture directed to pre-training a first multimodal transformer model on an unlabeled dataset, training a second multimodal transformer model on a first labeled dataset to perform a key information extraction task processing the unlabeled dataset with the second multimodal transformer model to generate pseudo-labels for the unlabeled dataset, training the first multimodal transformer model based on a second labeled dataset comprising one or more labels, the pseudo-labels generated, or combinations thereof to generate a third multimodal transformer model, generating updated pseudo-labels based on label completion predictions from the third multimodal transformer model, and training the third multimodal transformer model using a noise-aware loss function and the updated pseudo-labels to generate an updated third multimodal transformer model.Type: GrantFiled: April 10, 2023Date of Patent: January 2, 2024Assignee: Intuit, Inc.Inventors: Karelia Del Carmen Pena Pena, Tharathorn Rimchala, Peter Lee Frick, Tak Yiu Daniel Li
-
Patent number: 11810252Abstract: The present disclosure relates to methods, devices, and systems for blending geographic data when combining geographic data sources. The methods, devices, and systems identify a blend region for transitioning between a first dataset and a second dataset. The methods, devices, and systems extrapolate geographic data from the second dataset to blend with the geographic data from the first dataset to create blended elevation data in the blend region. The methods, devices, and systems may generate an image for a geographic region with the first set of geographic data, the second set of geographic data, and the blended elevation data.Type: GrantFiled: April 9, 2021Date of Patent: November 7, 2023Assignee: Microsoft Technology Licensing, LLCInventor: Duncan Murray Lawler
-
Patent number: 11734486Abstract: Aspects of the invention include systems and methods for implementing a sweepline triangulation technique to optimize spanning graphs for circuit routing. A non-limiting example computer-implemented method includes receiving an unrouted net having a plurality of elements. The elements can include pins, vias, and wires. A sweepline is passed across the unrouted net until the sweepline intersects an element of the plurality of elements. In response to the sweepline intersecting the element, the sweepline is stopped and one or more nodes on the sweepline and one or more previous nodes are identified. A connectivity graph is built from the one or more nodes and the one or more previous nodes. The connectivity graph includes one or more arcs and one or more guides. A minimum spanning tree is built by removing one or more guides from the connectivity graph and the unrouted net is routed based on the minimum spanning tree.Type: GrantFiled: September 7, 2021Date of Patent: August 22, 2023Assignee: International Business Machines CorporationInventors: Diwesh Pandey, Gustavo Enrique Tellez
-
Patent number: 11722682Abstract: A method of entropy coding in a video encoder is provided that includes assigning a first bin to a first single-probability bin encoder based on a probability state of the first bin, wherein the first single-probability bin encoder performs binary arithmetic coding based on a first fixed probability state, assigning a second bin to a second single-probability bin encoder based on a probability state of the second bin, wherein the second single-probability bin encoder performs binary arithmetic coding based on a second fixed probability state different from the first fixed probability state, and coding the first bin in the first single-probability bin encoder and the second bin in the second single-probability bin encoder in parallel, wherein the first single-probability bin encoder uses a first rLPS table for the first fixed probability state and the second single-probability bin encoder uses a second rLPS table for the second fixed probability state.Type: GrantFiled: December 30, 2020Date of Patent: August 8, 2023Assignee: Texas Instruments IncorporatedInventors: Vivienne Sze, Madhukar Budagavi
-
Patent number: 11709798Abstract: An example method is provided in according with one implementation of the present disclosure. The method comprises generating, via a processor, a set of hashes for each of a plurality of objects. The method also comprises computing, via the processor, a high-dimensional sparse vector for each object, where the vector represents the set of hashes for each object. The method further comprises computing, via the processor, a combined high-dimensional sparse vector from the high-dimensional sparse vectors for all objects and computing a hash suppression threshold. The method also comprises determining, via the processor, a group of hashes to be suppressed by using the hash suppression threshold, and suppressing, via the processor, the group of selected hashes when performing an action.Type: GrantFiled: October 13, 2021Date of Patent: July 25, 2023Assignee: Hewlett Packard Enterprise Development LPInventors: Mehran Kafai, Kave Eshghi, Omar Aguilar Macedo
-
Patent number: 11657540Abstract: A device and method used to image cylindrical fluid conduits, such as pipes, wellbores and tubulars, with ultrasound transducers then compress that data for storage or visualization. The compressed images may be stored on the tool and/or transmitted over telemetry, enabling the device to inspect and record long pipes or wells in high resolution on a single trip. This allow the ultrasound imaging tool to record much longer wells in higher resolution than would otherwise be possible. An outward-facing radial array of ultrasound transducers captures cross-sectional slices of the conduit to create frames from scan lines. The frames are compressed by applying a demodulation process and spatial conversion process to the scan lines. Video compression is applied to the to the demodulated, spatially converted ultrasound images to return compressed images.Type: GrantFiled: June 24, 2020Date of Patent: May 23, 2023Assignee: DarkVision Technologies IncInventor: Steven Wrinch
-
Patent number: 11657541Abstract: A method to compress an image includes assigning each pixel of the image to a cluster based on a red-green-blue (RGB) location of the pixel. The method also includes updating a centroid of the cluster after each pixel is assigned, based at least in part on the RGB location of the pixel, where the centroid is an RGB location. The method includes replacing each pixel in the image with an RGB value of the centroid of the cluster to which the pixel is assigned. The method also includes instructing a display to display a compressed image where, in the compressed image, each pixel in the image is replaced with the RGB value of the centroid of the cluster to which the pixel is assigned.Type: GrantFiled: December 31, 2020Date of Patent: May 23, 2023Assignee: TEXAS INSTRUMENTS INCORPORATEDInventors: Jeffrey Matthew Kempf, Jonathan Andrew Lucas
-
Patent number: 11601713Abstract: A system and method for identifying media segments using audio augmented image cross-comparison is disclosed, in which a media segment identifying system analyses both audio and video content, producing a unique identifier to compare with previously identified media segments in a media segment database. The characteristic landmark-linked-image-comparisons are constructed by first identifying an audio landmark. The audio landmark is an audio peak that exceeds a predetermined threshold. Two digital images are then obtained, one associated directly with the audio landmark, and one obtained a predetermined landmark time removed from the first image. The two images are then used to provide a characteristic landmark-linked-image-comparison. The pair of images are reduced in pixel size and converted to gray scale. Corresponding pixels are compared to form a numeric comparison. One image is mirrored before comparison to reduce the possibility of null comparisons.Type: GrantFiled: December 9, 2020Date of Patent: March 7, 2023Inventors: Oran Gilad, Samuel Chenillo, Oren Steinfeld
-
Patent number: 11587281Abstract: A graphics processor architecture provides for scan conversion and ray tracing approaches to visible surface determination as concurrent and separate processes. Surfaces can be identified for shading by scan conversion and ray tracing. Data produced by each can be normalized, so that instances of shaders, being executed on a unified shading computation resource, can shade surfaces originating from both ray tracing and rasterization. Such resource also may execute geometry shaders. The shaders can emit rays to be tested for intersection by the ray tracing process. Such shaders can complete, without waiting for those emitted rays to complete. Where scan conversion operates on tiles of 2-D screen pixels, the ray tracing can be tile aware, and controlled to prioritize testing of rays based on scan conversion status. Ray population can be controlled by feedback to any of scan conversion, and shading.Type: GrantFiled: January 11, 2021Date of Patent: February 21, 2023Assignee: Imagination Technologies LimitedInventors: John W. Howson, Luke Tilman Peterson, Steven J. Clohset
-
Patent number: 11582463Abstract: A method, computer program, and computer system is provided for aligning across layers in a coded video stream. A video bitstream having multiple layers is decoded. One or more subpicture regions are identified from among the multiple layers of the decoded video bitstream, the subpicture regions including a background region and one or more foreground subpicture regions. An enhanced subpicture is decoded and displayed based on a determination that a foreground subpicture region is selected. The background region is decoded and displayed based on a determination that a foreground subpicture region was not selected.Type: GrantFiled: October 5, 2020Date of Patent: February 14, 2023Assignee: TENCENT AMERICA LLCInventors: Byeongdoo Choi, Shan Liu, Stephan Wenger
-
Patent number: 11544943Abstract: A method includes executing an encoder machine learning model on multiple token values contained in a document to create an encoder hidden state vector. A decoder machine learning model executing on the encoder hidden state vector generates raw text comprising an entity value and an entity label for each of multiple entities. The method further includes generating a structural representation of the entities directly from the raw text and outputting the structural representation of the entities of the document.Type: GrantFiled: May 31, 2022Date of Patent: January 3, 2023Assignee: Intuit Inc.Inventors: Tharathorn Rimchala, Peter Frick
-
Patent number: 11501469Abstract: A data generation system for generating data representing content to be displayed includes: a content dividing unit operable to divide content to be displayed into a plurality of polyhedra and generate polyhedron position information, an intersection detecting unit operable to generate intersection information that describes the intersection of one or more surfaces within the content with the plurality of polyhedra, a polyhedron classifying unit operable to classify each of the polyhedra in dependence upon the intersection information, the classification indicating the properties of the surface within the respective polyhedra, and a data generating unit operable to generate data comprising the polyhedron position information and the polyhedron classification information.Type: GrantFiled: September 12, 2018Date of Patent: November 15, 2022Assignee: Sony Interactive Entertainment Inc.Inventor: Patrick John Connor
-
Patent number: 11475602Abstract: Initial low-quality images of a progressively-displayed high-definition image are masked with corresponding progressively-revealing mask filters or masking algorithms to realistically obscure such low quality and therefore to provide a realistically appearing progressive presentation of the high-definition image.Type: GrantFiled: October 31, 2020Date of Patent: October 18, 2022Assignee: PANAMORPH, INC.Inventor: Shawn L. Kelly
-
Patent number: 11477482Abstract: A three-dimensional data storage method includes: acquiring one or more units in which an encoded stream generated by encoding point cloud data is stored; and storing the one or more units into a file. The storing includes storing, in control information for the file, information indicating that data stored in the file is data generated by encoding the point cloud data.Type: GrantFiled: December 23, 2020Date of Patent: October 18, 2022Assignee: PANASONIC INTELLECTUAL PROPERTY CORPORATION OF AMERICAInventors: Noritaka Iguchi, Toshiyasu Sugio
-
Patent number: 11423867Abstract: Provided are a signal processing device and an image display apparatus including the same. The signal processing device and the image display apparatus including the same include a scaler configured to scale input images of various resolutions to a first resolution and a resolution enhancement processor configured to perform learning on the input images having the first resolution and to generate a first image having a second resolution higher than the first resolution. Accordingly, resolution may be stably improved even if input images of various resolutions are input.Type: GrantFiled: May 29, 2019Date of Patent: August 23, 2022Assignee: LG Electronics Inc.Inventor: Jungeun Lim
-
Patent number: 11405624Abstract: A moving image decoder (1) includes an intermediate estimated prediction mode deriving section (124) for transforming a prediction mode of each neighbor partition into an intermediate prediction mode included in an intermediate prediction set which is a sum of prediction sets (PS); and an estimated prediction mode deriving section (125) for deriving an estimated prediction mode by estimating a prediction mode of a target partition based on the intermediate prediction mode of each neighbor partition which is obtained by the transform.Type: GrantFiled: April 19, 2021Date of Patent: August 2, 2022Assignee: SHARP KABUSHIKI KAISHAInventors: Tomoyuki Yamamoto, Tomohiro Ikai
-
Patent number: 11397810Abstract: An information handling system improves removal of steganography data embedded in a graphics file by processing graphics files stored in a file system or transmitted through a network by processing the graphics files in a steganalyzer. The steganalyzer converts the body segment of the graphics file into binary code, and then compresses the binary code into a graphics file. This process results in the removal of any potential malicious code. The body segment location can be determined by parsing the portable network graphics file to determine a location of a pre-fix graphics file signature and a post-fix graphics file signature, with the graphics files signatures being specific to a particular type of graphics file.Type: GrantFiled: August 5, 2019Date of Patent: July 26, 2022Assignee: Dell Products L.P.Inventors: Yevgeni Gehtman, Maxim Futerman
-
Patent number: 11386597Abstract: An information processing apparatus (10) is for supporting work by a user who uses drawings for a plant. The information processing apparatus (10) includes a controller (15). The controller (15) is configured to generate an intermediate model, for at least one of a first drawing and a second drawing that include elements configuring the plant and are judged to have different formats, such that the format of the first drawing and the format of the second drawing are matched. The controller (15) is configured to judge whether a difference exists between the first drawing and the second drawing based on the generated intermediate model.Type: GrantFiled: December 25, 2020Date of Patent: July 12, 2022Assignee: YOKOGAWA ELECTRIC CORPORATIONInventors: Takahiro Kambe, Tatenobu Seki, Nobuaki Ema, Masato Annen
-
Patent number: 11348285Abstract: A method of compressing meshes using a projection-based approach, and leveraging the tools and syntax already generated for projection-based point cloud compression is described herein. Similar to the V-PCC approach, the mesh is segmented into surface patches, only the difference is that the segments follow the connectivity of the mesh. Each surface patch (or 3D patch) is then projected to a 2D patch, whereby in the case of the mesh, the triangle surface sampling is similar to a common rasterization approach used in computer graphics. For each patch, the position of the projected vertices is kept in a list, along with the connectivity of those vertices. The sampled surface now resembles a point cloud, and is coded with the same approach used for point cloud compression. Additionally, the list of vertices and connectivity is encoded per patch, and this data is sent along with the coded point cloud data.Type: GrantFiled: April 27, 2020Date of Patent: May 31, 2022Assignee: Sony Group CorporationInventors: Danillo Graziosi, Ohji Nakagami, Alexandre Zaghetto, Ali Tabatabai
-
Patent number: 11266480Abstract: Technology is described for augmenting medical imaging for use in a medical procedure. The method can include the operation of receiving an image of patient anatomy captured by a visual image camera during the medical procedure. An acquired medical image associated with the patient anatomy can then be retrieved. Another operation can be associating the acquired medical image to the patient anatomy. An augmentation tag associated with a location in one layer of the acquired medical image can be retrieved. A further operation can be projecting the acquired medical image and the augmentation tag using an augmented reality headset to form a single graphical view as an overlay to the patient anatomy in either 2D, 3D or holographic form.Type: GrantFiled: March 15, 2021Date of Patent: March 8, 2022Assignee: Novarad CorporationInventors: Wendell Arlen Gibby, Steven Todd Cvetko
-
Patent number: 11265542Abstract: Method for decoding a picture, comprising: decoding information that the picture is partitioned into more than one segment based on one or more syntax elements in a bitstream; decoding information that the spatial segmentation is uniform based on the one or more syntax elements; determining a segment unit size based on the one or more syntax elements or based on a predefined segment unit size; decoding a first value indicating a segment width from one or more code words in the bitstream; decoding a second value indicating a segment height from the one or more code words; deriving segment column widths based on a picture width in number of segment units and the first value; deriving segment row heights based on a picture height in number of segment units and the second value; deriving a spatial location for a current block based on the derived segment column widths and the derived segment heights; and decoding the current block based on the derived spatial location.Type: GrantFiled: March 5, 2021Date of Patent: March 1, 2022Assignee: TELEFONAKTIEBOLAGET LM ERICSSON (PUBL)Inventors: Mitra Damghanian, Martin Pettersson, Rickard Sjöberg
-
Patent number: 11122280Abstract: An image data of pictures constituting moving image data is encoded to generate an encoded video stream. In this case, the image data of the pictures constituting the moving image data is classified into a plurality of levels and encoded to generate a video stream having the image data of the pictures at the respective levels. Hierarchical composition is equalized between a low-level side and a high-level side, and corresponding pictures on the low-level side and the high-level side are combined into one set and are sequentially encoded. This allows a reception side to decode the encoded image data of the pictures on the low-level side and the high-level side with a smaller buffer size and a reduced decoding delay.Type: GrantFiled: August 23, 2019Date of Patent: September 14, 2021Assignee: SONY CORPORATIONInventor: Ikuo Tsukagoshi
-
Patent number: 11107462Abstract: Exemplary embodiments relate to improvements in spoken language understanding (SLU) systems. Conventionally, SLU systems include an automatic speech recognition (ASR) component configured to receive an input of audio data and to generate a textual representation of the audio data. Conventional SLU systems also include a natural language understanding (NLU) component configured to receive a text-based transcript and perform language-based tasks such as domain classification, intent determination, and slot-filling. However, these two components are typically trained separately based on different metrics. In real-world situations, errors in the ASR component propagate to the NLU component, which degrades the performance of the overall system. Exemplary embodiments described herein perform SLU in an end-to-end manner that infers semantic meaning directly from audio features without an intermediate text representation.Type: GrantFiled: October 30, 2018Date of Patent: August 31, 2021Assignee: FACEBOOK, INC.Inventors: Christian Fuegen, Yongquiang Wang, Anuj Kumar, Baiyang Liu, Dmitrii Serdiuk
-
Patent number: 11095901Abstract: A picture is obtained. The picture is intended to be compressed in a video stream. The picture includes pixel data that depicts a scene. A subject in the pixel data that depicts the scene is identified. The identification is based on the pixel data included in the picture. A plurality of object data is generated. The generated object data is related to the subject. The generation is based on the identification of the subject. An intended compression format of the pictured to be compressed in the video stream is determined. A pixel operation on the picture is performed. The performance is based on the determination of the intended compression format and before compression of the picture. The generated plurality of object data is related with the picture after compression of the picture.Type: GrantFiled: September 23, 2019Date of Patent: August 17, 2021Assignee: International Business Machines CorporationInventors: En-Shuo Hsu, Po-Hsun Tseng, David Shao Chung Chen, Wei-Te Chiang, Hsiao-Yung Chen
-
Patent number: 11089322Abstract: A method of decoding video data in merge mode can include constructing a merge candidate list using available spatial and temporal merge candidates; deriving motion information using a merge index and the merge candidate list; generating a prediction block using the motion information; generating a residual block by inverse-quantizing a quantized block using a quantization parameter and a quantization matrix and by inverse-transforming the inverse quantized block; and generating a reconstructed block using the residual block and the prediction block, wherein the quantization parameter is generated per quantization unit and a minimum size of the quantization unit is adjusted per picture by using a parameter which specifies the depth between the quantization unit having the minimum size and a largest coding unit, and the quantization parameter is generated using a quantization parameter predictor and a differential quantization parameter.Type: GrantFiled: December 12, 2018Date of Patent: August 10, 2021Assignee: INFOBRIDGE PTE. LTD.Inventors: Soo Mi Oh, Moonock Yang
-
Patent number: 11049240Abstract: The present invention relates to a method and system for assessing bone age using deep neural network, more specifically, in which regions of interest (ROIs) even for rotated objects can be more precisely and accurately extracted from an image by a rotated object detection technique used in region proposal networks. Thereby bones with different angles in the image can be detected with excellent speed and accuracy.Type: GrantFiled: May 23, 2019Date of Patent: June 29, 2021Assignee: HealthHub Co., Ltd.Inventors: Byung Il Lee, Sung Hyun Kim
-
Patent number: 11044478Abstract: A system comprises an encoder configured to compress images, such as image frames comprising attribute information and/or spatial for a point cloud and/or an occupancy map for the point cloud. Also, a system includes a decoder configured to decompress compressed image frames, such as image frames comprising compressed attribute and/or spatial information for the point cloud or an occupancy map for the point cloud. Additionally, the encoder may map N-bit data to M-bit code words, where M is less than N. Alternatively the encoder may map N-bit data to M-bit code words, where M is greater than N. In a similar manner, a decoder may map the M-bit code words back to the N-bit data.Type: GrantFiled: July 1, 2019Date of Patent: June 22, 2021Assignee: Apple Inc.Inventors: Alexandros Tourapis, Jungsun Kim, Fabrice A. Robinet, Khaled Mammou, Valery G. Valentin, Yeping Su
-
Patent number: 11037331Abstract: A method for de-contouring a source image is provided. In the method, a reference image is obtained based on the source image. Multi-oriented gradient calculation is performed on each pixel of the reference image, so as to obtain, for each pixel of the reference image, multiple gradient features that respectively correspond to multiple directions. For each pixel of the reference image, a monotonicity index is determined based on the corresponding gradient features. Then, a detail-protecting and de-contour operation is performed on the source image based on the monotonicity indices determined for the pixels of the reference image.Type: GrantFiled: January 21, 2020Date of Patent: June 15, 2021Assignee: NOVATEK MICROELECTRONICS CORP.Inventors: Cong Zhang, Jian-Hua Liang, Yuan-Jia Du
-
Patent number: 11032557Abstract: A moving image decoder (1) includes an intermediate estimated prediction mode deriving section (124) for transforming a prediction mode of each neighbor partition into an intermediate prediction mode included in an intermediate prediction set which is a sum of prediction sets (PS); and an estimated prediction mode deriving section (125) for deriving an estimated prediction mode by estimating a prediction mode of a target partition based on the intermediate prediction mode of each neighbor partition which is obtained by the transform.Type: GrantFiled: April 11, 2019Date of Patent: June 8, 2021Assignee: SHARP KABUSHIKI KAISHAInventors: Tomoyuki Yamamoto, Tomohiro Ikai
-
Patent number: 10984593Abstract: Techniques for designing and optimization of solid/cellular structures are described using a modeling process referred to as high-definition cellular level set in B-splines (HD-CLIBS). With this process, the entire design domain for the solid/cellular structure in question is subdivided into a set of connected volumetric cells in three dimensions. An implicit trivariate B-spline function is defined on each subdomain cell. With this parameterization scheme, constraints can be imposed on the relevant B-spline coefficients to naturally maintain geometric continuities at the connection faces between neighboring cells. The method offers several useful properties and powerful functionalities to build and modify a solid/cellular structure in the modeling process and to conduct topology optimization by directly adjusting the B-spline coefficients. The model construction can be carried out using a fast B-spline interpolation, and the topology optimization can involve a sequence of discrete B-spline convolutions.Type: GrantFiled: October 28, 2019Date of Patent: April 20, 2021Assignee: THE HONG KONG UNIVERSITY OF SCIENCE AND TECHNOLOGYInventor: Michael Yu Wang
-
Patent number: 10945807Abstract: Technology is described for augmenting medical imaging for use in a medical procedure. The method can include the operation of receiving an image of patient anatomy captured by a visual image camera during the medical procedure. An acquired medical image associated with the patient anatomy can then be retrieved. Another operation can be associating the acquired medical image to the patient anatomy. An augmentation tag associated with a location in one layer of the acquired medical image can be retrieved. A further operation can be projecting the acquired medical image and the augmentation tag using an augmented reality headset to form a single graphical view as an overlay to the patient anatomy in either 2D, 3D or holographic form.Type: GrantFiled: February 21, 2018Date of Patent: March 16, 2021Assignee: Novarad CorporationInventors: Wendell Arlen Gibby, Steven Todd Cvetko
-
Patent number: 10924784Abstract: A receiving side can perform interactive processing based on information of an object. Image data is coded to obtain a video stream having coded image data. The video stream is transmitted in a state of being added with information of an object detected on the basis of image data. For example, information of an object includes coded data obtained by coding one-bit data showing a shape of the object, information of a region that is a rectangular area enclosing the object, display priority information of the region, text information that explains the object, and the like. The receiving side can acquire information of an object without the need of detecting an object by processing image data, and without depending on its own performance, and is allowed to perform interactive processing based on information of an object in an excellent manner.Type: GrantFiled: August 17, 2017Date of Patent: February 16, 2021Assignee: SONY CORPORATIONInventor: Ikuo Tsukagoshi
-
Patent number: 10869059Abstract: A system comprises an encoder configured to compress a point cloud comprising a plurality of points each point comprising spatial information for the point. The encoder is configured to sub-sample the points and determine subdivision locations for the subsampled points. Also, the encoder is configured to determine, for respective subdivision location, if a point is to be included, not included, or relocated relative to the subdivision location. The encoder encodes spatial information for the sub-sampled points and encodes subdivision location point inclusion/relocation information to generate a compressed point cloud. A decoder recreates an original or near replica of an original point cloud based on the spatial information and the subdivision location inclusion/relocation information included in the compressed point cloud.Type: GrantFiled: May 11, 2020Date of Patent: December 15, 2020Assignee: Apple Inc.Inventors: Khaled Mammou, Fabrice A. Robinet, Andrea Cremaschi, Alexandros Tourapis
-
Patent number: 10856001Abstract: A polygon unit-based image processing method, and a device for the same are disclosed. Specifically, a method for decoding an image on the basis of a polygon unit can comprise the steps of: deriving a motion vector predictor for a polygon apex forming the polygon unit; deriving a motion vector for the polygon apex on the basis of a motion vector difference for the polygon apex and the motion vector predictor; and deriving a prediction sample for the polygon unit from a division unit, which is specified by the motion vector, in a reference picture.Type: GrantFiled: February 15, 2016Date of Patent: December 1, 2020Assignee: LG ELECTRONICS INC.Inventors: Moonmo Koo, Sehoon Yea, Eunyong Son, Jin Heo
-
Patent number: 10778970Abstract: A method of decoding video data using quantized coefficient components and inter prediction information can include extracting quantized coefficient components and inter prediction information from a received bit stream; applying an inverse scan pattern to the quantized coefficient components to generate a quantized block having a size of a transform unit; generating a quantization parameter per quantization unit which is a unit for deriving a quantization parameter and inverse-quantizing the quantized block to generate a transformed block; generating a residual block by inverse-transforming the transformed block; deriving motion information and generating a prediction block; and generating a reconstructed block by using the residual block and the prediction block.Type: GrantFiled: January 8, 2019Date of Patent: September 15, 2020Assignee: INFOBRIDGE PTD. LTD.Inventors: Soo Mi Oh, Moonock Yang
-
Patent number: 10762407Abstract: A component incorporating a 3-D identification code includes: a component body having an interior bounded by an exterior surface; and an identification code formed as a part of at least one of the interior and the exterior surface, the identification code including a plurality of cells arranged in a three-dimensional space, wherein each of the cells is configured to encode more than two possible values.Type: GrantFiled: April 5, 2017Date of Patent: September 1, 2020Assignee: General Electric CompanyInventor: Scott Alan Gold
-
Patent number: 10764470Abstract: A system and method for tile type based color space transformation improves the performance and image quality of multifunction devices. The disclosed embodiments transform an input image on a tile by tile basis. If a tile is detected as neutral, simple 1D L* to CMYK Tone Reproduction Curves or Look-Up-Tables are used to convert the input pixels of the tile in L*a*b* to output pixels in CMYK. If a tile is detected as containing color content, then the input pixels are chrominance adjusted and subsequently converted to CMYK using regular tetrahedral interpolation.Type: GrantFiled: March 11, 2019Date of Patent: September 1, 2020Assignee: Xerox CorporationInventors: Xing Li, Peter McCandlish, Clara Cuciurean-Zapan
-
Patent number: 10706529Abstract: Digital image processing device and method for holistic evaluation of subtle irregularities in a digital image by using the scale space technique to identify irregularities of interest and by calculating a total irregularity score using a function of intensity, scale and optionally location of the identified irregularities of interest. Specifically, the digital image represents a liquid mixture formed by mixing two or more liquid compositions of different ingredients, colors, viscosities, and/or solubility; the subtle irregularities represent non-homogenous mixing spots or regions in such liquid mixture.Type: GrantFiled: June 7, 2018Date of Patent: July 7, 2020Assignee: The Procter & Gamble CompanyInventors: Fabio Zonfrilli, Qi Zhang
-
Patent number: 10659816Abstract: A system comprises an encoder configured to compress a point cloud comprising a plurality of points each point comprising spatial information for the point. The encoder is configured to sub-sample the points and determine subdivision locations for the subsampled points. Also, the encoder is configured to determine, for respective subdivision location, if a point is to be included, not included, or relocated relative to the subdivision location. The encoder encodes spatial information for the sub-sampled points and encodes subdivision location point inclusion/relocation information to generate a compressed point cloud. A decoder recreates an original or near replica of an original point cloud based on the spatial information and the subdivision location inclusion/relocation information included in the compressed point cloud.Type: GrantFiled: September 12, 2019Date of Patent: May 19, 2020Assignee: Apple Inc.Inventors: Khaled Mammou, Fabrice A. Robinet, Andrea Cremaschi, Alexandros Tourapis
-
Patent number: 10606887Abstract: The present disclosure is directed towards methods and systems for providing relevant video scenes in response to a video search query. The systems and methods identify a plurality of key frames of a media object and detect one or more content features represented in the plurality of key frames. Based on the one or more detect content features, the systems and methods associate tags indicating the detected content features with the plurality of key frames of the media object. The systems and methods, in response to receiving a search query including search terms, compare the search terms with the tags of the selected key frames, identify a selected key frame that depicts at least one content feature related to the search terms, and provide a preview image of the media item depicting the at least one content feature.Type: GrantFiled: September 23, 2016Date of Patent: March 31, 2020Assignee: Adobe Inc.Inventors: Sachin Soni, Ashish Duggal, Anmol Dhawan
-
Patent number: 10515269Abstract: A method is described that includes receiving raw image data corresponding to a series of raw images, and processing the raw image data with an encoder of a processing device to generate encoded data. The encoder is characterized by an input/output transformation that substantially mimics the input/output transformation of at least one retinal cell of a vertebrate retina. The method also includes processing the encoded data to generate dimension reduced encoded data by applying a dimension reduction algorithm to the encoded data. The dimension reduction algorithm is configured to compress an amount of information contained in the encoded data. An apparatus and system usable with such a method is also described.Type: GrantFiled: April 20, 2016Date of Patent: December 24, 2019Assignee: CORNELL UNIVERSITYInventor: Sheila Nirenberg
-
Patent number: 10499068Abstract: There is provided methods, apparatuses and computer program products for video coding and decoding. A first part of a first coded video representation is decoded, and information on decoding a second coded video representation is received and parsed. The coded second representation differs from the first coded video representation in chroma format, sample bit depth, color gamut and/or spatial resolution, and the information indicates if the second coded video representation may be decoded using processed decoded pictures of the first coded video representation as reference pictures.Type: GrantFiled: December 29, 2015Date of Patent: December 3, 2019Assignee: Nokia Technologies OyInventor: Miska Matias Hannuksela
-
Patent number: 10451560Abstract: According to one implementation, a damage detection system includes optical paths, a light source, a photodetector, and a signal processing system, a signal processing system. The optical paths propagate lights in at least three different directions. The optical paths have at least two paths per one direction. The light source makes the lights incident on one ends of the optical paths respectively. The photodetector detects the lights output from other ends of the optical paths. The signal processing system specifies at least one location of damage based on optical detection signals detected by the photodetector.Type: GrantFiled: January 25, 2019Date of Patent: October 22, 2019Assignee: SUBARU CORPORATIONInventor: Masakatsu Abe
-
Patent number: 10438631Abstract: A method for real time video processing for retouching an object in a video is presented The method includes providing an object in the video steam, where the object is at least partially and at least occasionally presented in frames of the video. The method sets a degree of retouching and generates a list of at least one element of the object selected based on a request of retouching and the degree of retouching. The method detects the at least one element of the object in the video and parameters of the at least one element and calculates new parameters of the at least one element according to the degree of retouching. Characteristic points are detected for each of the at least one element of the object and a mesh is generated based on the characteristic points for each of the at least one element of the object. The at least one element of the object in the video is tracked by aligning the mesh for each of the at least one element with a position of the corresponding each of the at least one element.Type: GrantFiled: June 25, 2014Date of Patent: October 8, 2019Assignee: Snap Inc.Inventors: Elena Shaburova, Victor Shaburov
-
Patent number: 10341671Abstract: An image compression method includes compressing an input image with first and second compression methods to generate first and second (e.g., lossless and lossy) compressed images. First and second residual layers are generated, based on a difference between the first and second compressed images. Connected components in the residual layers are identified. Each connected component includes a group of one or more pixels that, when mapped to the second compressed image is connected, in first and second directions, to pixels in the second compressed image. A compressed image is generated, which includes, for a connected component identified in the first residual layer, removing corresponding pixels from the second compressed image, and for a connected component identified in the second residual layer, adding corresponding pixels to the second compressed image. The system and method can thus provide file size savings associated with lossy compression while avoiding character replacement.Type: GrantFiled: April 21, 2017Date of Patent: July 2, 2019Assignee: XEROX CorporationInventors: Sainarayanan Gopalakrishnan, Xing Li, Rajasekar Kanagasabai, Clara Cuciurean-Zapan, Vignesh Doss
-
Patent number: 10313332Abstract: A method of performing one-time password (OTP) authentication using a color code is provided. The method includes generating, by an OTP terminal and an authentication server, an OTP using a reference time, generating, by the OTP terminal, a color code corresponding to a binary code representing the OTP, displaying, by a display device, the color code, obtaining, by a camera, the displayed color code, decoding, by the authentication server, the obtained color code into the color code and generating a password which refers to the decoded binary code, and performing, by the authentication server, authentication by comparing the generated OTP and the password.Type: GrantFiled: August 30, 2016Date of Patent: June 4, 2019Assignee: Research & Business Foundation Sungkyunkwan UniversityInventors: Kwang-Seok Hong, Sang Min Park
-
Patent number: 10311564Abstract: An image processing device is provided with a storage unit configured to store for each model among a plurality of models, a model template obtained by reducing a model image; an image reduction unit configured to reduce an input image by the reduction ratio for each model; and a search unit configured to search for the location of a model in an input image on the basis of a reduced input image and a model template. The storage unit stores model templates for at least two of the models wherein said model templates are obtained by reducing the model images by the same reduction ratio; and The image reduction unit processes the at least two models jointly. The image processing device is hereby capable of high-speed identification and recognition of a plurality of models from an input image.Type: GrantFiled: March 14, 2014Date of Patent: June 4, 2019Assignee: OMRON CorporationInventor: Yoshinori Konishi
-
Patent number: 10313689Abstract: A moving image decoder (1) includes an intermediate estimated prediction mode deriving section (124) for transforming a prediction mode of each neighbor partition into an intermediate prediction mode included in an intermediate prediction set which is a sum of prediction sets (PS); and an estimated prediction mode deriving section (125) for deriving an estimated prediction mode by estimating a prediction mode of a target partition based on the intermediate prediction mode of each neighbor partition which is obtained by the transform.Type: GrantFiled: February 15, 2018Date of Patent: June 4, 2019Assignee: SHARP KABUSHIKI KAISHAInventors: Tomoyuki Yamamoto, Tomohiro Ikai
-
Patent number: 10277806Abstract: A method comprising using at least one hardware processor for: receiving a continuous stream of images; and in real time, for each image of some of the images: extracting composition features from the image, determining photographic composition functions for at least some of the extracted composition features based on a plurality of composition rules, defining a photographic composition sum-function by summing up at least some of the photographic composition functions, and calculating an extremum of the photographic composition sum-function, wherein the extremum reflects a photographic composition suggestion with respect to a photographic composition of the image.Type: GrantFiled: June 28, 2015Date of Patent: April 30, 2019Assignee: ARTINCAM LTD.Inventors: Ofri Masad, Ariel Shamir