专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明专利

AU2021247504A1 Method and measurement system for determining characteristics of particles of a bulk material 未知
公开(公告)号：AU2021247504A1
公开(公告)日：2022-09-08
申请号：AU2021247504
申请日：2021-03-31
申请人： KVERNELAND GROUP NIEUW VENNEP B V
发明人： DE BOER BART
IPC分类号： G06T7/62 , A01C7/10 , G06K9/00 , G06T7/80 , G06V10/24
摘要： The present disclosure refers to a method for determining characteristics of particles of a bulk material such as fertilizer, seed or the like, comprising: providing a heap of particles (2a) of a bulk material (2) to be distributed such as fertilizer, seed or the like by a distribution machine; providing a measurement tool (1) having an optical landmark (7) on a front side of the measurement tool (1); providing the measurement tool (1) in a measurement position in which the heaped particles (2a) of the bulk material (2) are provided in proximity to the measurement tool (1); providing a camera device (3) configured to detect images; detecting image data by the camera device (3), the image data indicative of an image of the front side of the measurement tool (1) provided in the measurement position and the heaped particles (2a) of the bulk material (2) provided in proximity to the measurement tool (1); and determining characteristics of the particles (2a) of the bulk material (2) from image data analysis of the image data. The determining, in one or more processors, comprises: determining the optical landmark (7) from the image data; in response to determining the optical landmark (7), providing dimensional data assigned to and indicative of dimensional characteristics of the optical landmark (7); and determining characteristics of the particles (2a) of the bulk material (2) taking into account the dimensional characteristics. Further, a measurement system for determining characteristics of particles of a bulk material is provided.

2. 发明专利

AU2022203118A1 Computer-implemented method for extracting content from a physical writing surface 未知
公开(公告)号：AU2022203118A1
公开(公告)日：2022-06-16
申请号：AU2022203118
申请日：2022-05-10
申请人： INKERZ PTY LTD
发明人： KOLAHDOUZAN VAHID , AMINAIEE ABDOLHOSSEIN , KOLAHDOUZAN MASOUD
IPC分类号： G06V10/22 , G06T7/13 , G06T7/181 , G06T7/33 , G06V10/24 , G06V10/26 , G06V30/10 , G06V30/14 , G06V30/148 , G06V30/414 , G06V30/418
摘要： COMPUTER-IMPLEMENTED METHOD FOR EXTRACTING CONTENT FROM A PHYSICAL WRITING SURFACE Abstract A computer-implemented method (300) for extracting content (302) from a physical writing surface (304), the method (300) comprising the steps of: (a) receiving a reference frame (306) including image data relating to at least a portion of the physical writing surface (304), the image data including a set of data points; (b) determining an extraction region (308), the extraction region (308) including a subset of the set of data points from which content (302) is to be extracted; (c) extracting content (302) from the extraction region (308) and writing the content (302) to a display frame (394); (d) receiving a subsequent frame (406) including subsequent image data relating to at least a portion of the physical writing surface (304), the subsequent image data including a subsequent set of data points; (e) determining a subsequent extraction region (408), the subsequent extraction region (408) including a subset of the subsequent set of data points from which content (402) is to be extracted; and (f) extracting subsequent content (402) from the subsequent extraction region (408) and writing the subsequent content (402) to the display frame (394). Network120 .of Devic~s)22 Irierfa201) 214 o[mniuio1 Spcil Fnco Fig.2A

3. 发明专利

GB2601940A Methods, systems and apparatus for segmenting and dimensioning objects 未知
公开(公告)号：GB2601940A
公开(公告)日：2022-06-15
申请号：GB202202633
申请日：2017-06-16
申请人： SYMBOL TECHNOLOGIES LLC
发明人： YE GU , YAN ZHANG
IPC分类号： G06T7/33 , G06F17/16 , G06T7/80 , G06V10/24
摘要： Disclosed is a method of generating a transformation matrix that transforms the data from a first coordination system of a first image sensor to a second coordination of a second image sensor. The method starts by generating an initial transformation matrix based on a two dimensional image data representation of an area 1002. A refined matrix is generated from the initial matrix and a three dimensional image data representation of the area, by aligning depth values of the 3D image data according to the initial matrix 1004, extracting overlapping points in the aligned depth values1008, discarding non-overlapping points in the aligned depth values 1010, and refining the initial matrix according to the overlapping points 1012. The refining may be based on offsets between the overlapping points, the refining of th initial matrix according to the overlapping points includes generating multipliers to apply to the initial matrix based on the offsets between the overlapping points.

4. 发明专利

GB2603697A Determining object orientation from an image with machine learning 未知
公开(公告)号：GB2603697A
公开(公告)日：2022-08-10
申请号：GB202205313
申请日：2020-10-26
申请人： NVIDIA CORP
发明人： JONATHAN TREMBLAY , MING-YU LIU , DIETER FOX , PHILIP AMMIRATO
IPC分类号： G06V10/24
摘要： Apparatuses, systems, and techniques to determine orientation of an objects in an image. In at least one embodiment, images are processed using a neural network trained to determine orientation of an object.

5. 发明专利

AU2020374849A1 Determining object orientation from an image with machine learning 未知
公开(公告)号：AU2020374849A1
公开(公告)日：2022-06-23
申请号：AU2020374849
申请日：2020-10-26
申请人： NVIDIA CORP
发明人： TREMBLAY JONATHAN , LIU MING-YU , FOX DIETER , AMMIRATO PHILIP
IPC分类号： G06V10/24 , G06V30/146
摘要： Apparatuses, systems, and techniques to determine orientation of an objects in an image. In at least one embodiment, images are processed using a neural network trained to determine orientation of an object.

6. 发明专利

GB2600587A Image defect identification 未知
公开(公告)号：GB2600587A
公开(公告)日：2022-05-04
申请号：GB202200865
申请日：2020-06-12
申请人： IBM
发明人： FAN LI , GUO QIANG HU , SHENG NAN ZHU , JUN ZHU , JING CHANG HUANG , PENG JI , YUAN YUAN DING
IPC分类号： G06K9/62 , G06V10/24
摘要： A method, a device and a computer program product for image processing are proposed. In the method, whether a first image indicates a defect associated with a target object is determined. In response to determining that the first image indicates the defect, a second image absent from the defect is obtained based on the first image. The defect is identified by comparing the first image with the second image. In this way, the defect associated with the target object in the image can be accurately and efficiently identified or segmented.

7. 实用新型

CN220962462U 一种下相机自动对位装置有权
公开(公告)号：CN220962462U
公开(公告)日：2024-05-14
申请号：CN202322411116.1
申请日：2023-09-05
申请人：禾洛半导体(徐州)有限公司
发明人：黃思閎 , 鄭仁迪
IPC分类号： G06V10/14 , G06V10/24
摘要：本实用新型涉及一种下相机自动对位装置，包括相机组件和光源组件，其特征在于：还包括调节轴和相机衬套，相机衬套套住相机组件并与之固定，调节轴通过小皮带轮、皮带和大皮带轮转动连接相机衬套，小皮带轮套住调节轴并与之固定，大皮带轮套住相机衬套并与之固定，调节轴依次通过小皮带轮、皮带和大皮带轮带动相机衬套和相机组件同步转动；光源组件设置在相机组件的正上方。本实用新型通过相机衬套，将相机灵活安装在相机安装支架中，在通过调节轴、小齿轮、皮带和大齿轮带动相机衬套和相机组件同步转动，从而可快速、方便、精准的调节相机组件的取像角度，不仅通用性强，适用性广，而且后期维护或更换都比较方便。

8. 发明专利

NL2028357B1 METHOD FOR TRAINING A NEURAL NETWORK 未知
公开(公告)号：NL2028357B1
公开(公告)日：2022-12-13
申请号：NL2028357
申请日：2021-06-01
申请人： CYCLOMEDIA TECH B V
发明人： JULIEN ADRIANUS VIJVERBERG , BART JOHANNES BEERS , BASTIAAN JOHANNES BOOM
IPC分类号： G06V10/24 , G06V10/774 , G06V10/82 , G06V20/56
摘要： The present invention relates to a method for training a neural network. The present invention further relates to a method for training a neural network for recognizing an object, to a method for detecting an object in an image using a neural network, and to a method for detecting an object in an image. The present invention proposes to use a first and second camera that are mounted on a vehicle and that have the same orientation. The first and second cameras are different cameras. By triggering the first and second camera such that at a time of recording an image, the cameras are at substantially the same position, a dataset is obtained that is used for training a neural network for transforming an image recorded by the first camera into an image that would have been recorded by the second camera.

9. 发明专利

GB2602513A Method and apparatus for monitoring objects using computer vision 未知
公开(公告)号：GB2602513A
公开(公告)日：2022-07-06
申请号：GB202101370
申请日：2021-02-01
申请人： SEECHANGE TECH LIMITED
发明人： DAVID PACKWOOD , ARIEL EDGAR RUIZ GARCIA , ELEFTHERIOS FANIOUDAKIS
IPC分类号： G06V10/24 , G06N3/02 , G06V10/40 , G06V10/70
摘要： The method and apparatus classify and localise objects to determine when an object has passed through a predefined space. The method comprises: receiving image data (102 fig.1) captured by a camera 360, localising and classifying objects (104, fig.1) using a trained multi-headed system comprising a single classifier head paired with a single localiser head. Location and classification data are generated and used to determine that an object 322 has traversed a space 330 when an object is classified and localised in two different predetermined spatial and/or temporal locations 330a,b,c. The multi-headed classifier may comprise a convolutional neural network (CNN) backbone that outputs processed data to the single classification and localisation heads which may comprise fully connected neural networks. A traversal may be identified by determining that an object has been detected in two or more detection zones defining horizontally adjacent physical spaces within the captured images. The object classification may be used to count the number of each class of object that has traversed the monitored space. The system may be used to monitor objects being passed through a space at a checkout, or at the entrance/exit of a car park.

10. 发明专利

AR121227A1 SISTEMAS Y MÉTODOS PARA ANÁLISIS CONTEXTUAL DE IMÁGENES 未知
公开(公告)号：AR121227A1
公开(公告)日：2022-04-27
申请号：ARP210100247
申请日：2021-01-29
申请人： COSMO ARTIFICIAL INTELLIGENCE AI LTD
发明人： CHERUBINI ANDREA , NGO DINH NHAN
IPC分类号： G06V10/24 , G06V30/146
摘要： En una implementación, se proporciona un sistema implementado por computadora para el procesamiento de video en tiempo real. El sistema incluye por lo menos una memoria configurada para almacenar instrucciones y por lo menos un procesador configurado para ejecutar las instrucciones para realizar operaciones. El por lo menos un procesador está configurado para recibir video en tiempo real generado por un sistema de imágenes médicas, donde el video en tiempo real incluye una pluralidad de cuadros de imagen, y obtener información de contexto que indica una interacción de un usuario con el sistema de imágenes médicas. El por lo menos procesador también está configurado para realizar una detección de objetos para detectar por lo menos un objeto en la pluralidad de cuadros de imagen, y realizar una clasificación para generar información de clasificación para por lo menos un objeto en la pluralidad de cuadros de imagen. Además, el por lo menos un procesador está configurado para realizar una manipulación de video a fin de modificar el video en tiempo real recibido, sobre la base de por lo menos una de la detección de objetos y la clasificación. Además, el procesador está configurado para invocar por lo menos una de la detección de objetos, la clasificación y la manipulación de video sobre la base de la información de contexto.

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