专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

71. 发明专利

FR3087038B1 PROCEDE DE SUIVI D'ELEMENTS D'INTERET VISIBLES DANS UNE VIDEO 未知
公开(公告)号：FR3087038B1
公开(公告)日：2021-03-19
申请号：FR1859158
申请日：2018-10-03
申请人： IDEMIA IDENTITY & SECURITY FRANCE
发明人： JOURDAS CECILE , CSILLAG DORA , THIEBAUT MAXIME
IPC分类号： G06T7/246 , G06N3/08
摘要： La présente invention concerne un procédé de suivi d'éléments d'intérêt visibles dans une vidéo constituée d'une séquence de K images, caractérisé en ce qu'il comprend la mise en œuvre par des moyens de traitement de données (21) d'un terminal (2), d'étapes de : (a) Association de chaque élément d'intérêt d'une première catégorie visible dans une k-ième image de ladite vidéo avec un élément d'intérêt d'une deuxième catégorie différente de la première catégorie visible dans ladite k-ième image ; (b) Calcul d'un coût d'association d'une pluralité de paires d'un élément d'intérêt de la première catégorie visible dans au moins une image de la vidéo avec un élément d'intérêt de la deuxième catégorie visible dans au moins une image de la vidéo, en fonction d'au moins des compteurs d'association des paires d'un élément d'intérêt de la première catégorie avec un élément d'intérêt de la deuxième catégorie ; (c) Mise en œuvre d'un algorithme d'optimisation combinatoire en fonction des coûts d'association calculés de sorte à réassocier chaque élément d'intérêt de la première catégorie visible dans ladite k-ième image avec un élément d'intérêt de la deuxième catégorie visible dans ladite k-ième image ; (d) Mise à jour des compteurs d'association.

72. 发明专利

AU2019311751A1 Image turbulence correction using tile approach 未知
公开(公告)号：AU2019311751A1
公开(公告)日：2021-03-11
申请号：AU2019311751
申请日：2019-04-24
申请人： RAYTHEON CO
发明人： HAYES MATTHEW JOHN FRANCIS , BEARDSLEY CHRISTOPHER J , FLUCKIGER DAVID U
IPC分类号： G06T3/00 , G06T7/246
摘要： System and method for image turbulence correction includes: receiving a plurality of consecutive image frames; demosaicing previous, current and preceding image frames into a plurality of same size overlapping video tiles; determining a displacement of each of the video tiles; converting the video tiles of the current image frame, the previous image frame, and the plurality of preceding image frames into a frequency domain; iteratively processing the video tiles of the previous image frame, the current image frame and the plurality of preceding image frames for turbulence correction in the frequency domain; converting the turbulence corrected video tiles into a spatial domain, wherein the converted turbulence corrected video tiles form a single video frame tile including turbulence degradation correction; and mosaicing the single video frame tiles including turbulence degradation correction together to generate a full field of view turbulence correct video stream.

73. 发明专利

GB2567920B Deep salient content neural networks for efficient digital object segmentation 未知
公开(公告)号：GB2567920B
公开(公告)日：2021-02-17
申请号：GB201813276
申请日：2018-08-15
申请人： ADOBE INC
发明人： ZHE LIN , XIN LU , XIAOHUI SHEN , JIMEI YANG , JIANMING ZHANG , JEN-CHAN JEFF CHIEN , CHENXI LIU
IPC分类号： G06T7/10 , G06K9/32 , G06K9/46 , G06K9/62 , G06N3/02 , G06T7/246
摘要： Systems, methods, and non-transitory computer-readable media are disclosed for segmenting objects in digital visual media utilizing one or more salient content neural networks. In particular, in one or more embodiments, the disclosed systems and methods train one or more salient content neural networks to efficiently identify foreground pixels in digital visual media. Moreover, in one or more embodiments, the disclosed systems and methods provide a trained salient content neural network to a mobile device, allowing the mobile device to directly select salient objects in digital visual media utilizing a trained neural network. Furthermore, in one or more embodiments, the disclosed systems and methods train and provide multiple salient content neural networks, such that mobile devices can identify objects in real-time digital visual media feeds (utilizing a first salient content neural network) and identify objects in static digital images (utilizing a second salient content neural network).

74. 发明专利

CA3141956A1 LAYERED MOTION REPRESENTATION AND EXTRACTION IN MONOCULAR STILL CAMERA VIDEOS 未知
公开(公告)号：CA3141956A1
公开(公告)日：2021-01-07
申请号：CA3141956
申请日：2020-05-18
申请人： EVERSEEN LTD
发明人： GUI VASILE , PESCARU DAN , CERNAZANU-GLAVAN COSMIN
IPC分类号： G06T7/215 , G06T7/11 , G06T7/143 , G06T7/174 , G06T7/194 , G06T7/246 , G06T7/254 , G06T7/277
摘要： A method for segmenting motion data and inferring layering order in a video includes segmenting a first image frame of the video using intra-frame mode estimation, segmenting further image frames of the video using inter-frame mode matching, generating a label change map for a current segmented image frame by generating a binary map of the current segmented image frame, and then marking one or more pixels whose labels have changed with respect to a previous segmented image frame, detecting one or more non-still image segments of the current segmented image frame, detecting one or more moving image segments in the one or more non-still image segments, estimating motion of the one or more moving image segments, representing each moving segment as a vector of seven parameters, and clustering one or more similar moving segment vectors based on corresponding one or more parameters.

75. 发明专利

CA3143817A1 STICKER GENERATING METHOD AND APPARATUS, AND MEDIUM AND ELECTRONIC DEVICE 未知
公开(公告)号：CA3143817A1
公开(公告)日：2020-12-30
申请号：CA3143817
申请日：2020-05-22
申请人： BEIJING BYTEDANCE NETWORK TECH CO LTD
发明人： WEI YANTONG
IPC分类号： G06T7/246
摘要： A sticker generating method, comprising: obtaining a background image, the background image comprising a target object (S101); displaying a display area of a sticker and anchors of the sticker in the background image (S102); receiving an import instruction of the sticker (S103); importing resource of the sticker according to the import instruction and displaying the resource of the sticker in the display area of the sticker (S104); dynamically selecting a tracking area according to the positions of the anchors of the sticker, wherein the tracking area is an image area of the target object (S105); and generating the sticker according to the display area, the tracking area, and the resource of the sticker (S106). By means of the sticker generating method, when a sticker moves on a target object, an image area of the target object can be selected in real time, a relative position relationship between the sticker and the target object can be determined more accurately, the target object can be tracked more precisely, the position of the sticker is accurately set, the position of the sticker can be dynamically adjusted according to a target object detection result, and a fast-responsive real-time dynamic sticker effect is achieved.

76. 发明专利

ES2798759T3 Formación de imágenes de flujo y mediciones basadas en fluorescencia 未知
公开(公告)号：ES2798759T3
公开(公告)日：2020-12-14
申请号：ES16821893
申请日：2016-07-06
申请人： SCINOVIA CORP
发明人： SUND JAMES , COHEN DAVID
IPC分类号： G06T7/246 , A61B5/0275 , G06T7/00 , G06T7/254
摘要： Un método de seguimiento basado en fluorescencia de un marcador (606) emisor de luz en una corriente (620) de fluido corporal, el método comprende: monitorizar, con una cámara (610), una región de interés atravesada por la corriente (620) de fluido corporal en la que se ha introducido un marcador (606) emisor de luz; grabar datos de video en movimiento generados por la cámara (610); determinar características de tiempo de los datos de video en movimiento; y calcular características de flujo en base a las características de tiempo; el método comprende adicionalmente: dividir (202) los datos de video en movimiento en núcleos; identificar (204) cual de los núcleos recibe alguna porción del marcador emisor de luz utilizando un umbral de intensidad; computar (206), para cada núcleo identificado, un conjunto de datos de señal de intensidad que comprende información de intensidad de luz media versus tiempo; realizar suavizamiento (208) en cada conjunto de datos de señal de intensidad; calcular (210) un tiempo de retardo entre los conjuntos de datos de señal de intensidad de núcleos identificados vecinos utilizando correlación cruzada; utilizar una resolución espacial y el tiempo de retardo, calcular (212) vectores de velocidad; sumar (214) los vectores de velocidad de los núcleos vecinos para crear un vector de velocidad resultante; y generar (216) un mapa de velocidad de los vectores de velocidad resultantes para todos los núcleos.

77. 发明专利

AU2019275236A1 System and method for sensor pointing control 未知
公开(公告)号：AU2019275236A1
公开(公告)日：2020-07-30
申请号：AU2019275236
申请日：2019-04-25
申请人： RAYTHEON CO
发明人： VERA RICHARD C , JOHNSTON COLIN M , CANTY APRIL DANBEE
IPC分类号： G06T7/246 , G06T7/73 , H04N5/232
摘要： A system and method for pointing an imaging sensor. In one embodiment, the method includes forming a correlation surface; calculating, from the correlation surface, a first estimated target offset, the first estimated target offset being an estimate of an offset between a view of a target in the first subimage and a view of the target in the second subimage; and adjusting a pointing angle of the imaging sensor according to the first estimated target offset.

78. 发明专利

GB2579736A Image processing device and image processing method 未知
公开(公告)号：GB2579736A
公开(公告)日：2020-07-01
申请号：GB202002056
申请日：2017-08-22
申请人： MITSUBISHI ELECTRIC CORP
发明人： YOSHIMI MORIYA , NAOHIRO SHIBUYA , KAZUO SUGIMOTO
IPC分类号： G06T7/246 , G06T7/292
摘要： This image processing device is provided with: an object detection unit (103a) which detects objects appearing in image data; an object tracking unit (103b) which tracks the detected objects along a time axis; a descriptor generation unit (104) which generates feature descriptors for the objects on the basis of the object detection results and the object tracking results; a data recording control unit (105) which constructs a database from the generated feature descriptors and image data of the objects associated with the feature descriptors; an image search unit (108a) which searches the database for a pair of objects, each of which meets search criteria and appears in image data obtained by capturing an image of a different region; a feature extraction unit (108b) which extracts feature quantities used for matching purposes, from the image data and feature descriptors for the found pair of objects; and a determination unit (108c) which determines whether or not the pair of objects are identical to each other by comparing the extracted feature quantities.

79. 发明专利

GB2579634A Method and apparatus for determining temporal behaviour of an object 未知
公开(公告)号：GB2579634A
公开(公告)日：2020-07-01
申请号：GB201820003
申请日：2018-12-07
申请人： PHASE FOCUS LTD
发明人： KEVIN LANGLEY , MATTHEW STAGG , JAMES RUSSELL
IPC分类号： G06T7/246
摘要： A computer-implemented method and apparatus for determining the temporal behaviour of an object are disclosed. The method comprises receiving image track data 505 indicative of a plurality of tracks. Each track identifies an association between features corresponding to an object in each of a plurality of images. In at least some 510 of the tracks, one or more faults are identified 515, with each error associated with at least one corresponding fault image from the data. A correction process is performed iteratively (550-580, Fig. 5b) on the track data. The rectification procedure comprises selecting a defective track 540 and identifying one or more candidate features 545 in each respective fault image, where each potential feature is determined as a candidate for correcting at least one error. One or more candidate corrections are determined for the selected track, and the best correction is selected 555 in dependence on one or more criteria. The technique may proceed in a loop. and may evaluate whether different potential corrections are mutually inconsistent. It may be used to track cells.

80. 发明专利

DE102019115241A1 VORRICHTUNG UND VERFAHREN ZUM ERFASSEN EINER LANGSAMEN FAHRZEUGBEWEGUNG 未知
公开(公告)号：DE102019115241A1
公开(公告)日：2020-05-07
申请号：DE102019115241
申请日：2019-06-05
申请人： GM GLOBAL TECH OPERATIONS LLC
发明人： MEHDI SYED B , HU YASEN
IPC分类号： G06T7/246 , G08G1/16
摘要： Es werden ein Verfahren und eine Vorrichtung zum Erkennen einer Gefährdung durch ein Fahrzeug bereitgestellt. Das Verfahren beinhaltet das Erfassen eines Fahrzeugrades in einer Vielzahl von Einzelbildern eines Videos, das Erzeugen von Begrenzungsrahmen um Abschnitte der Einzelbilder, einschließlich des Fahrzeugrades, das Skalieren der Abschnitte der Einzelbilder, einschließlich des Fahrzeugrades, auf eine vorbestimmte konstante Größe, die Bestimmung, ob sich das Fahrzeugrad bewegt, durch Analysieren der skalierten Abschnitte des Bildes, und das Ausgeben von Informationen, die anzeigen, dass sich das Fahrzeug bewegt, wenn die Bestimmung ergibt, dass sich das Fahrzeugrad bewegt.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式