公开(公告)号：US07616816B2

公开(公告)日：2009-11-10

申请号：US11688433

申请日：2007-03-20

Applicant: Hui Cheng

Inventor： Hui Cheng

IPC: G06K9/00

CPC classification number: G06F17/30241 ,  G06F17/30831 ,  G06F17/30849 ,  Y10S707/99933

Abstract: The present invention provides a virtual data collection system and method that provides an image analyst with full-control of the image, including the ability to retrieve and reorganize information from an existing video database of images. The system provides for the search and retrieval of information from previously collected images, and reorganize them to meet new or revised mission objectives. The system also enables quick access of the “right” information through virtual data collection from the existing video database.

Abstract translation: 本发明提供一种虚拟数据收集系统和方法，为图像分析者提供对图像的全面控制，包括从图像的现有视频数据库检索和重组信息的能力。 该系统提供从以前收集的图像中搜索和检索信息，并重新组织以满足新的或修订的任务目标。 该系统还可以通过现有视频数据库的虚拟数据收集快速访问“正确”信息。

公开(公告)号：US20070273696A1

公开(公告)日：2007-11-29

申请号：US11695755

申请日：2007-04-03

Applicant: Hui Cheng ,  Darren Butler

Inventor： Hui Cheng ,  Darren Butler

IPC: G06T11/00

CPC classification number: G06T7/00 ,  G06K9/00771 ,  G06K9/469 ,  G06T2207/10016 ,  G06T2207/20072

Abstract: A method for transforming Video-To-Text is disclosed that automatically generates text descriptions of the content of a video. The present invention first segments an input video sequence according to predefined semantic classes using a Mixture-of-Experts blob segmentation algorithm. The resulting segmentation is coerced into a semantic concept graph and based on domain knowledge and a semantic concept hierarchy. Then, the initial semantic concept graph is summarized and pruned. Finally, according to the summarized semantic concept graph and its changes over time, text and/or speech descriptions are automatically generated using one of the three description schemes: key-frame, key-object and key-change descriptions.

Abstract translation: 公开了一种用于转换视频到文本的方法，其自动生成视频内容的文本描述。 本发明首先使用专家混合分组分割算法根据预定义的语义类别划分输入视频序列。 所得到的分割被强制转化为语义概念图，并且基于领域知识和语义概念层次。 然后，总结和修剪初始语义概念图。 最后，根据总结的语义概念图及其随时间的变化，使用三个描述方案之一自动生成文本和/或语音描述：关键帧，关键对象和键改变描述。

公开(公告)号：US20070092122A1

公开(公告)日：2007-04-26

申请号：US11521109

申请日：2006-09-14

Applicant: Jiangjian Xiao ,  Hui Cheng

Inventor： Jiangjian Xiao ,  Hui Cheng

IPC: G06K9/00

CPC classification number: H04N19/577 ,  G06T7/207 ,  G06T7/215 ,  H04N19/186 ,  H04N19/537 ,  H04N19/543 ,  H04N19/553 ,  H04N19/61

Abstract: The methods and systems of the present invention enable the estimation of optical flow by performing color segmentation and adaptive bilateral filtering to regularize the flow field to achieve a more accurate flow field estimation. After creating pyramid models for two input image frames, color segmentation is performed. Next, starting from a top level of the pyramid, additive flow vectors are iteratively estimated between the reference frames by a process including occlusion detection, wherein the symmetric property of backward and forward flow is enforced for the non-occluded regions. Next, a final estimated optical flow field is generated by expanding the current pyramid level to the next lower level and the repeating the process until the lowest level is reached. This approach not only generates efficient spatial-coherent flow fields, but also accurately locates flow discontinuities along the motion boundaries.

Abstract translation: 本发明的方法和系统能够通过执行颜色分割和自适应双边滤波来估计光流，以使流场正规化以实现更准确的流场估计。 在为两个输入图像帧创建金字塔模型之后，执行颜色分割。 接下来，从金字塔的顶级开始，通过包括遮挡检测的处理在参考帧之间迭代地估计加性流矢量，其中对于未遮挡区域强制执行向后和向前流的对称属性。 接下来，通过将当前金字塔等级扩展到下一个较低级别并重复该过程直到达到最低级别来生成最终估计的光流场。 这种方法不仅产生有效的空间相干流场，而且还准确地定位沿着运动边界的流动不连续性。

公开(公告)号：US20060245502A1

公开(公告)日：2006-11-02

申请号：US11401990

申请日：2006-04-10

Applicant: Hui Cheng ,  Michael Isnardi ,  Arkady Kopansky

Inventor： Hui Cheng ,  Michael Isnardi ,  Arkady Kopansky

IPC: H04N11/04

CPC classification number: H04N19/59 ,  H04N19/117 ,  H04N19/132 ,  H04N19/139 ,  H04N19/147 ,  H04N19/176 ,  H04N19/196 ,  H04N19/61 ,  H04N19/82 ,  H04N19/86

Abstract: A system and method of compressing a video signal can include the steps of: receiving a video signal, the video signal including frames; analyzing, for each frame, the video signal on a macroblock-by-macroblock level; determining whether to downsample a macroblock residual for each of the macroblocks; selectively downsampling a macroblock residual for some of the macroblocks; and coding the macroblocks. A system and method of decompressing a video signal can include the steps of receiving a compressed video signal, the video signal including frames; analyzing, for each frame, the video signal on a macroblock-by-macroblock level; determining whether to upsample a macroblock residual for each of the macroblocks; selectively upsampling a macroblock residual for some of the macroblocks; and decoding the macroblocks.

公开(公告)号：US07006256B2

公开(公告)日：2006-02-28

申请号：US09683540

申请日：2002-01-16

Applicant: Shen-ge Wang ,  Hui Cheng ,  Zhigang Fan ,  Ying-wei Lin

Inventor： Shen-ge Wang ,  Hui Cheng ,  Zhigang Fan ,  Ying-wei Lin

IPC: H04N1/46

CPC classification number: G06T1/0028 ,  G06T2201/0051 ,  G06T2201/0061 ,  H04N1/32203 ,  H04N1/32208 ,  H04N1/32256 ,  H04N2201/3233 ,  H04N2201/327

Abstract: A watermarked image generator includes a watermark data source that inputs watermark data to a watermark embedding device. The watermark embedding device halftones the input image to generate the output image made of 2×2 binary patterns, the 2×2 binary patterns forming the watermarks embedded in the output image. The watermark embedding device includes a tri-level error diffusion generator that performs tri-level error diffusion on the input image by halftoning the input image into black, white and 50% gray. A plurality of halftoning circuits successively replace each pixel of the halftoned input image with one of a plurality of the 2×2 binary patterns. The one of the plurality of 2×2 binary patterns corresponds to at least one respective bit of the watermark data when a pixel of the halftoned image data is 50% gray. The watermark data can be segmentation map data of the input image.

公开(公告)号：US06871036B2

公开(公告)日：2005-03-22

申请号：US10624558

申请日：2003-07-23

Applicant: Kenji Asuwa ,  Toru Miyasaka ,  Tomio Sugaya ,  Hiroyuki Mabuchi ,  Hui Cheng

Inventor： Kenji Asuwa ,  Toru Miyasaka ,  Tomio Sugaya ,  Hiroyuki Mabuchi ,  Hui Cheng

IPC: B65H5/02 ,  G03G15/00 ,  G03G15/01

CPC classification number: G03G15/755

Abstract: A belt driving apparatus which inhibits a belt meandering so as to improve a printing quality and rapidly starts a printing motion at a time of starting the apparatus such as a time of turning on power or the like, and an image forming apparatus using the same. The belt driving apparatus has an endless belt, a plurality of belt tension supports for supporting the belt in a tension manner, and a steering roller supported in a tiltable manner and compensating a meandering of the belt, and the belt driving apparatus is provided with a meandering compensation sensitivity adjuster for adjusting a belt meandering compensation sensitivity of the steering roller per an angle of incline.

Abstract translation: 一种带驱动装置，其抑制带曲折以提高打印质量，并且在启动诸如打开电源等的时间之类的装置时快速开始打印动作，以及使用该打印质量的图像形成装置。 带驱动装置具有环形带，用于以张力方式支撑带的多个带张力支撑件和以可倾斜的方式支撑并且补偿带的曲折的转向辊，并且带驱动装置设置有 蜿蜒补偿灵敏度调节器，用于每倾斜角度调整转向辊的带曲折补偿灵敏度。

公开(公告)号：US06695435B1

公开(公告)日：2004-02-24

申请号：US10448135

申请日：2003-05-30

Applicant: Hui Cheng ,  Beilei Xu ,  Elliott A. Eklund

Inventor： Hui Cheng ,  Beilei Xu ,  Elliott A. Eklund

IPC: B41J221

CPC classification number: B41J2/2139 ,  B41J2/0451 ,  B41J2/04586 ,  H04N1/58

Abstract: Selective replacement methods and systems based on a stochastic process use a redundancy inherent, for example, in a four-plus color printing system that has full-width colors. Since the absorption spectra of the four or more colors usually overlap to some extent, there is a redundancy that can be used to reduce artifacts, such as streaks. Since the human visual system is much less sensitive to chrominance changes, than to luminance changes, matching the luminance of a pixel where a color is missing, to the luminance of the original color or to the luminance of the pixels that neighbor that pixel can adequately to reduce the severity of the artifact caused by the missing color while reducing the chance that the replacing process itself introduces any additional artifacts.

Abstract translation: 基于随机过程的选择性替换方法和系统使用固有的冗余，例如在具有全宽颜色的四色彩色打印系统中。 由于四种或更多种颜色的吸收光谱通常在一定程度上重叠，所以存在可用于减少伪像（例如条纹）的冗余。 由于人类视觉系统对于色度变化的敏感程度要比亮度变化（匹配缺失颜色的像素的亮度）与原始颜色的亮度或邻近该像素的像素的亮度相匹配 以减少由缺失颜色引起的伪像的严重性，同时减少替换过程本身引入任何其他工件的机会。

公开(公告)号：US08913783B2

公开(公告)日：2014-12-16

申请号：US12913861

申请日：2010-10-28

Applicant: Saad Masood Khan ,  Hui Cheng ,  Dennis Lee Matthies ,  Harpreet Singh Sawhney ,  Sang-Hack Jung ,  Chris Broaddus ,  Bogdan Calin Mihai Matei ,  Ajay Divakaran

Inventor： Saad Masood Khan ,  Hui Cheng ,  Dennis Lee Matthies ,  Harpreet Singh Sawhney ,  Sang-Hack Jung ,  Chris Broaddus ,  Bogdan Calin Mihai Matei ,  Ajay Divakaran

IPC: G06K9/00 ,  G06K9/46

CPC classification number: G06K9/00785 ,  B64C39/024 ,  B64C2201/123 ,  B64D47/08 ,  G06K9/00651 ,  G06K9/4671 ,  G06K9/6269 ,  G06K9/6282 ,  G06T7/11 ,  G06T7/20 ,  G06T2207/10012

Abstract: A computer implemented method for determining a vehicle type of a vehicle detected in an image is disclosed. An image having a detected vehicle is received. A number of vehicle models having salient feature points is projected on the detected vehicle. A first set of features derived from each of the salient feature locations of the vehicle models is compared to a second set of features derived from corresponding salient feature locations of the detected vehicle to form a set of positive match scores (p-scores) and a set of negative match scores (n-scores). The detected vehicle is classified as one of the vehicle models models based at least in part on the set of p-scores and the set of n-scores.

Abstract translation: 公开了一种用于确定在图像中检测到的车辆的车辆类型的计算机实现的方法。 接收具有检测到的车辆的图像。 在检测到的车辆上投影出具有显着特征点的多个车辆模型。 将来自车辆模型的每个突出特征位置的第一组特征与从所检测到的车辆的相应突出特征位置导出的第二组特征进行比较，以形成一组正的匹配得分（p分数）和 一组负面比赛得分（n分）。 检测到的车辆至少部分地基于p分数集合和n分数集合被分类为车辆模型模型之一。

公开(公告)号：US08712096B2

公开(公告)日：2014-04-29

申请号：US13041092

申请日：2011-03-04

Applicant: Jiangjian Xiao ,  Harpreet Singh Sawhney ,  Hui Cheng

Inventor： Jiangjian Xiao ,  Harpreet Singh Sawhney ,  Hui Cheng

IPC: G06K9/00

CPC classification number: G06K9/6296 ,  G06K9/00624 ,  G06T7/254

Abstract: The present invention relates to a method and apparatus for detecting and tracking vehicles. One embodiment of a system for detecting and tracking an object (e.g., vehicle) in a field of view includes a moving object indication stage for detecting a candidate object in a series of input video frames depicting the field of view and a track association stage that uses a joint probabilistic graph matching framework to associate an existing track with the candidate object.

Abstract translation: 本发明涉及车辆检测跟踪方法和装置。 用于在视野中检测和跟踪对象（例如，车辆）的系统的一个实施例包括用于检测描绘视场的一系列输入视频帧中的候选对象的移动对象指示级和跟踪关联级， 使用联合概率图匹配框架将现有轨道与候选对象相关联。

公开(公告)号：US08634638B2

公开(公告)日：2014-01-21

申请号：US12488911

申请日：2009-06-22

Applicant: Feng Han ,  Hui Cheng ,  Jiangjian Xiao ,  Harpreet Singh Sawhney ,  Sang-Hack Jung ,  Rakesh Kumar ,  Yanlin Guo

Inventor： Feng Han ,  Hui Cheng ,  Jiangjian Xiao ,  Harpreet Singh Sawhney ,  Sang-Hack Jung ,  Rakesh Kumar ,  Yanlin Guo

IPC: G06K9/62

CPC classification number: G06K9/00342

Abstract: The present invention relates to a method and system for creating a strong classifier based on motion patterns wherein the strong classifier may be used to determine an action being performed by a body in motion. When creating the strong classifier, action classification is performed by measuring similarities between features within motion patterns. Embodiments of the present invention may utilize candidate part-based action sets and training samples to train one or more weak classifiers that are then used to create a strong classifier.

Abstract translation: 本发明涉及一种用于基于运动模式创建强分类器的方法和系统，其中强分类器可用于确定由运动中的身体执行的动作。 当创建强分类器时，通过测量运动模式中的特征之间的相似度来执行动作分类。 本发明的实施例可以利用基于候选部分的动作集和训练样本来训练一个或多个弱分类器，然后将其用于创建强分类器。