-
公开(公告)号：US20190095730A1
公开(公告)日：2019-03-28
申请号：US16136128
申请日：2018-09-19
Inventor： Huiyuan FU , Huadong MA , Yifan ZHANG , Yu CAO , Dengkui ZHANG
Abstract: Embodiments of the present invention provide an end-to-end lightweight method and apparatus for license plate recognition. The method comprises: obtaining an image to be recognized; obtaining a number of a license plate in the image to be recognized and position coordinates of the license plate in the image to be recognized on the basis of the image to be recognized and a pre-trained target license plate recognition model, wherein the target license plate recognition model comprises a target feature extraction network, a target region candidate localization network, a target super-resolution generation network and a target recurrent neural network. Because in this solution, once an image to be recognized is input into the target license plate recognition model, the target license plate recognition model can output the license plate number and position coordinates of the license plate in the image to be recognized, one realizes an end-to-end model. The model has relatively strong robustness, and it can detect and recognize pictures taken under different camera angles. Moreover, computation variables such as image features can be reused without repeated computations, the model takes up less RAM and the speed of license plate recognition is greatly improved.
-
公开(公告)号：US20230067605A1
公开(公告)日：2023-03-02
申请号：US17456581
申请日：2021-11-25
Inventor： Liang LIU , Xiaolong ZHENG , Huadong MA , Chengling JIANG , Zihui LUO
Abstract: A deep reinforcement learning (DRL)-based intelligent job batching method and apparatus, and an electronic device are provided. The method includes: obtaining static features and a dynamic feature of each job, where the static features of the job include a delivery date, a specification and a process requirement of the job, and the dynamic feature of the job includes a receiving moment; and inputting the static features and the dynamic feature of each job into a job batching module, and using a Markov decision process (MDP) by the job batching module to combine jobs with similar features in a to-be-batched job set into an identical batch, so as to minimize a total quantity of batches obtained finally and a difference in features of jobs in each batch. The DRL-based intelligent job batching method and apparatus can learn a stable batching strategy and provide a stable and efficient job batching solution.
-
Patent Agency Ranking
公开(公告)号：US20200279560A1
公开(公告)日：2020-09-03
申请号：US16675816
申请日：2019-11-06
Inventor： Huanpu YIN , Anfu ZHOU , Liang LIU , Huadong MA
Abstract: Embodiments of the present application provide an acoustic sensing-based text input method, comprising: obtaining audio information corresponding to text to be input; dividing the audio information to obtain an audio segment for each letter to be recognized in the text to be input; sending to the server, a type of the text to be input, the audio segments for letters to be recognized, and arrangement of the audio segment for the letter to be recognized in the audio information; receiving input result returned by the server, and displaying, based on the input result, text information corresponding to the text to be input on the display screen of the mobile terminal, The method allows effective text input without relying on a display screen.
-
4.发明申请公开(公告)号：US20160334546A1
公开(公告)日：2016-11-17
申请号：US14845859
申请日：2015-09-04
Inventor： Huadong MA , Huiyuan FU , Zheng ZHANG
CPC classification number: G01W1/00 , G06K9/00664 , G06K9/4642 , G06K9/6269
Abstract: The embodiments of the present invention disclose a weather recognition method and device based on image information detection, which includes: obtaining an image to be detected; extracting multiple first image features of the image to be detected with respect to each preset type of weather according to a number of first preset algorithms preset correspondingly for different preset types of weather; inputting the extracted multiple first image features to a preset multi-kernel classifier, the multi-kernel classifier performing classification according to the inputted image features to identify the weather in which the image to be detected was taken; wherein the multi-kernel classifier is a classifier for the preset types of weather realized by: selecting a first preset number of image samples for the different preset types of weather in which the image was taken respectively; and for the image samples of each preset type of weather respectively, extracting the first image features of each image sample according to the first preset algorithm which corresponds to this preset type of weather; and performing machine learning for the extracted first image features according to a preset multi-kernel learning algorithm. The weather in which the image was taken can be identified by applying the solutions provided by the embodiments of the present invention.
Abstract translation: 本发明的实施例公开了一种基于图像信息检测的天气识别方法和装置，其包括：获得要检测的图像; 根据针对不同预设类型的天气预先相应地设定的第一预设算法的数量，提取相对于每种预设类型的天气的要检测的图像的多个第一图像特征; 将所提取的多个第一图像特征输入到预设的多内核分类器中，所述多内核分类器根据所输入的图像特征进行分类，以识别拍摄所述图像的天气; 其中所述多内核分类器是通过以下方式实现的预设类型的天气的分类器：为分别拍摄图像的不同预设类型的天气选择第一预设数量的图像样本; 并分别针对每种预设类型的天气的图像样本，根据与该预设类型的天气相对应的第一预设算法提取每个图像样本的第一图像特征; 并根据预设的多核学习算法对所提取的第一图像特征执行机器学习。 拍摄图像的天气可以通过应用本发明实施例提供的解决方案来识别。
-
公开(公告)号：US20200343968A1
公开(公告)日：2020-10-29
申请号：US16856721
申请日：2020-04-23
Inventor： Liang LIU , Huadong MA , Pengrui DUAN , Dan XIA
IPC: H04B7/185 , H04W84/06 , H04L12/721 , H04L12/751 , H04L12/715 , H04L12/717 , H04L12/741 , H04W40/24
Abstract: A routing method and apparatus for SDN-based LEO satellite network are disclosed. The LEO satellite network includes a control plane and a data plane. The control plane includes a central controller and a plurality of local controllers. The data plane includes a plurality of LEO satellite nodes and user terminals connecting to the LEO satellite nodes. The control plane may be located on the earth, and thus the centralized management and control of the data plane are placed on the earth. A local controllers monitors LEO satellite nodes in a subnet or subnets of the local controller. The distance between a local controller and a LEO satellite node is much smaller than the distance between a GEO satellite node and the LEO satellite node, and thus the time delay and the traffic loss of communication are reduced.
-
公开(公告)号：US20180060684A1
公开(公告)日：2018-03-01
申请号：US15350813
申请日：2016-11-14
Inventor： Huadong MA , Wu LIU , Xinchen LIU , Haitao ZHANG , Huiyuan FU
CPC classification number: G06K9/325 , G06F17/30253 , G06F17/30256 , G06K9/00771 , G06K9/6215
Abstract: The present application discloses a vehicle searching method and device, which can perform the steps of: calculating an appearance similarity distance between a first image of a target vehicle and several second images containing the searched vehicle; selecting several images from the several second images as several third images; obtaining corresponding license plate features of license plate areas in the first image and each of the third images with a preset Siamese neural network model; calculating a license plate feature similarity distance between the first image and each of the third images according to license plate feature; calculating a visual similarity distance between the first image and each of the third images according to the appearance similarity distance and the license plate feature similarity distance; obtaining a the first search result of the target vehicle by arranging the several third images in an ascending order of the visual similarity distances. The solution provided by the present application is not limited by application scenes, and it also improves vehicle searching speed and accuracy while reducing requirements of hardware such as cameras that collect images of a vehicle and auxiliary devices.
-
7.发明申请公开(公告)号：US20160337640A1
公开(公告)日：2016-11-17
申请号：US15012442
申请日：2016-02-01
Inventor： Huadong MA , Liang LIU , Aner LIU , Dawei LU
CPC classification number: H04N13/296 , G06T7/13 , G06T7/60 , H04N13/275
Abstract: A method and a system for determining parameters of an off-axis virtual camera provided by embodiments of present invention can extract a scene depth map for each video frame from a depth buffer, determine the minimum value of edge depth values of the scene depth map as the closest scene edge depth of each video frame, determine the depth of a first object as the depth of an object of interest of each video frame, use the smaller value between the closest scene edge depth and the depth of an object of interest as the zero-parallax value and obtain a zero-parallax value sequence constituted by the zero-parallax value of each video frame. The present invention realizes automatic determination of the zero parallax of each video frame rather than manual setting thereof, and thus the determination will not be affected by factors such as lack of experience, and the amount of work for an technician is also reduced. Meanwhile, the present invention makes the zero parallax as close as possible to an object of interest without incurring the window occlusion problem, which can improve the stereo perception as much as possible while ensuring the comfortability, and can ensure the using effect of the determined zero parallax.
Abstract translation: 本发明实施例提供的用于确定离轴虚拟摄像机的参数的方法和系统可以从深度缓冲器中提取每个视频帧的场景深度图，将场景深度图的边缘深度值的最小值确定为 每个视频帧的最接近的场景边缘深度，确定第一个对象的深度作为每个视频帧的感兴趣对象的深度，使用最近的场景边缘深度和感兴趣对象的深度之间的较小值作为 零视差值，并获得由每个视频帧的零视差值构成的零视差值序列。 本发明实现了每个视频帧的零视差的自动确定，而不是其手动设置，因此该决定不会受到诸如经验不足等因素的影响，并且技术人员的工作量也减少。 同时，本发明使零视差尽可能接近感兴趣的对象，而不会产生窗口遮挡问题，这可以在确保舒适性的同时尽可能地改善立体感觉，并且可以确保确定的零的使用效果 视差。
-
-
-
-
-
-
Search Results
7
records
(took 0.015 seconds)
