知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

73 records (took 0.049 seconds)

    Trellis based word decoder with reverse pass
    65.
    发明授权
    Trellis based word decoder with reverse pass 有权

    公开(公告)号：US09639783B2

    公开(公告)日：2017-05-02

    申请号：US14698528

    申请日：2015-04-28

    Applicant: QUALCOMM Incorporated

    Inventor： Pawan Kumar Baheti ,  Kishor K. Barman ,  Raj Kumar Krishna Kumar

    IPC: G06K9/34 ,  G06K9/72

    CPC classification number: G06K9/72 ,  G06K2209/01

    Abstract: Systems, apparatuses, and methods to relate images of words to a list of words are provided. A trellis based word decoder analyses a set of OCR characters and probabilities using a forward pass across a forward trellis and a reverse pass across a reverse trellis. Multiple paths may result, however, the most likely path from the trellises has the highest probability with valid links. A valid link is determined from the trellis by some dictionary word traversing the link. The most likely path is compared with a list of words to find the word closest to the most.

    Method and apparatus for low complexity compression of signals employing differential operation for transient segment detection
    66.
    发明授权
    Method and apparatus for low complexity compression of signals employing differential operation for transient segment detection 有权
    用于低复杂度压缩信号的方法和装置，采用差分操作进行瞬态段检测

    公开(公告)号：US09356731B2

    公开(公告)日：2016-05-31

    申请号：US14634117

    申请日：2015-02-27

    Applicant: QUALCOMM Incorporated

    Inventor： Pawan Kumar Baheti ,  Harinath Garudadri ,  Yuejie Chi

    IPC: G10L19/00 ,  G10L21/00 ,  H04L1/00 ,  H03M7/30 ,  H04L1/22

    CPC classification number: H04L1/0043 ,  H03M7/30 ,  H04L1/0052 ,  H04L1/22

    Abstract: Certain aspects of the present disclosure relate to techniques for low-complexity encoding (compression) of broad class of signals, which are typically not well modeled as sparse signals in either time-domain or frequency-domain. First, the signal can be split in time-segments that may be either sparse in time domain or sparse in frequency domain, for example by using absolute second order differential operator on the input signal. Next, different encoding strategies can be applied for each of these time-segments depending in which domain the sparsity is present.

    Abstract translation: 本公开的某些方面涉及广泛类别的信号的低复杂度编码（压缩）的技术，其通常不能被良好地建模为时域或频域中的稀疏信号。 首先，信号可以在时域中被分割，时间段可能在时域中稀疏或频域稀疏，例如通过在输入信号上使用绝对二阶微分算子。 接下来，不同的编码策略可以应用于这些时间段中的每一个，这取决于稀疏在哪个域中。

    Parameter selection and coarse localization of interest regions for MSER processing
    67.
    发明授权
    Parameter selection and coarse localization of interest regions for MSER processing 有权
    MSER处理的兴趣区域的参数选择和粗定位

    公开(公告)号：US09183458B2

    公开(公告)日：2015-11-10

    申请号：US13796729

    申请日：2013-03-12

    Applicant: QUALCOMM Incorporated

    Inventor： Pawan Kumar Baheti ,  Kishor K. Barman ,  Dhananjay Ashok Gore ,  Senthilkumar Sundaram

    IPC: G06K9/46 ,  G06K9/32

    CPC classification number: G06K9/4661 ,  G06K9/3233 ,  G06K9/3258 ,  G06K9/4638 ,  G06K9/4642 ,  G06K2209/01

    Abstract: An attribute is computed based on pixel intensities in an image of the real world, and thereafter used to identify at least one input for processing the image to identify at least a first maximally stable extremal region (MSER) therein. The at least one input is one of (A) a parameter used in MSER processing or (B) a portion of the image to be subject to MSER processing. The attribute may be a variance of pixel intensities, or computed from a histogram of pixel intensities. The attribute may be used with a look-up table, to identify parameter(s) used in MSER processing. The attribute may be a stroke width of a second MSER of a subsampled version of the image. The attribute may be used in checking whether a portion of the image satisfies a predetermined test, and if so including the portion in a region to be subject to MSER processing.

    Abstract translation: 基于现实世界的图像中的像素强度来计算属性，然后用于识别用于处理图像的至少一个输入以识别其中的至少第一最大稳定的极值区域（MSER）。 至少一个输入是（A）在MSER处理中使用的参数或（B）要进行MSER处理的图像的一部分中的一个。 该属性可以是像素强度的方差，或者从像素强度的直方图计算。 该属性可以与查找表一起使用，以识别在MSER处理中使用的参数。 属性可以是图像的子采样版本的第二MSER的笔画宽度。 该属性可以用于检查图像的一部分是否满足预定的测试，并且如果包括要进行MSER处理的区域中的部分。

    PROCESSING TEXT IMAGES WITH SHADOWS
    68.
    发明申请
    PROCESSING TEXT IMAGES WITH SHADOWS 有权
    用SHADOWS处理文本图像

    公开(公告)号：US20150193667A1

    公开(公告)日：2015-07-09

    申请号：US14150682

    申请日：2014-01-08

    Applicant: QUALCOMM Incorporated

    Inventor： Hemanth P. Acharya ,  Pawan Kumar Baheti ,  Kishor K. Barman

    IPC: G06K9/18 ,  G06T5/10 ,  G06T5/00 ,  G06T5/40

    CPC classification number: G06K9/18 ,  G06K9/38 ,  G06T7/10 ,  G06T2207/20021 ,  G06T2207/30244

    Abstract: Embodiments disclosed facilitate robust, accurate, and reliable recovery of words and/or characters in the presence of non-uniform lighting and/or shadows. In some embodiments, a method to recover text from image may comprise: expanding a Maximally Stable Extremal Region (MSER) in an image, the neighborhood comprising a plurality of sub-blocks; thresholding a subset of the plurality of sub-blocks in the neighborhood, the subset comprising sub-blocks with text, wherein each sub-block in the subset is thresholded using a corresponding threshold associated with the sub-block; and obtaining a thresholded neighborhood.

    Abstract translation: 所公开的实施例有助于在存在非均匀照明和/或阴影的情况下对字和/或字符的鲁棒，准确和可靠的恢复。 在一些实施例中，从图像中恢复文本的方法可以包括：在图像中扩展最大稳定的极大区域（MSER），所述邻域包括多个子块; 阈值化所述邻域中的所述多个子块的子集，所述子集包括具有文本的子块，其中使用与所述子块相关联的对应阈值对所述子集中的每个子块进行阈值化; 并获得阈值邻域。

    Automatic correction of skew in natural images and video
    69.
    发明授权
    Automatic correction of skew in natural images and video 有权
    自动修正自然图像和视频中的偏斜

    公开(公告)号：US09076242B2

    公开(公告)日：2015-07-07

    申请号：US13831237

    申请日：2013-03-14

    Applicant: QUALCOMM Incorporated

    Inventor： Pawan Kumar Baheti ,  Kishor K. Barman ,  Hemanth P. Acharya

    IPC: H04N5/232 ,  G06K9/36 ,  G06T3/60 ,  G06K9/32

    CPC classification number: G06T3/608 ,  G06K9/325 ,  G06K9/3283 ,  G06K2209/01 ,  H04N5/23222

    Abstract: An electronic device and method use a camera to capture an image of an environment outside followed by identification of regions therein. A subset of the regions is selected, based on attributes of the regions, such as aspect ratio, height, and variance in stroke width. Next, a number of angles that are candidates for use as skew of the image are determined (e.g. one angle is selected for each region. based on peakiness of a histogram of the region, evaluated at different angles). Then, an angle that is most common among these candidates is identified as the angle of skew of the image. The just-described identification of skew angle is performed prior to classification of any region as text or non-text. After skew identification, at least all regions in the subset are rotated by negative of the skew angle, to obtain skew-corrected regions for use in optical character recognition.

    Abstract translation: 电子设备和方法使用相机来捕获外部环境的图像，然后识别其中的区域。 基于区域的属性，例如纵横比，高度和笔画宽度方差，选择区域的子集。 接下来，确定作为图像的偏斜使用的候选者的多个角度（例如，基于区域的直方图的峰值，以不同的角度进行评估，针对每个区域选择一个角度）。 然后，在这些候选中最常见的角度被识别为图像的偏斜角。 在将任何区域分类为文本或非文本之前执行刚刚描述的倾斜角的识别。 在偏斜识别之后，子集中的至少所有区域以歪斜角度的相位旋转，以获得用于光学字符识别的偏斜校正区域。

    Identifying regions of text to merge in a natural image or video frame
    70.
    发明授权
    Identifying regions of text to merge in a natural image or video frame 有权
    识别要在自然图像或视频帧中合并的文本区域

    公开(公告)号：US09053361B2

    公开(公告)日：2015-06-09

    申请号：US13748539

    申请日：2013-01-23

    Applicant: QUALCOMM Incorporated

    Inventor： Pawan Kumar Baheti ,  Dhananjay Ashok Gore ,  Hyung-Il Koo ,  Te-Won Lee

    IPC: G06K9/00 ,  G06T11/60 ,  G06K9/36 ,  G06K9/32

    CPC classification number: G06K9/00456 ,  G06K9/3258 ,  G06K9/36 ,  G06K9/4647 ,  G06K2209/01 ,  G06T11/60

    Abstract: In several aspects of described embodiments, an electronic device and method use a camera to capture an image or a frame of video of an environment outside the electronic device followed by identification of blocks of regions in the image. Each block that contains a region is checked, as to whether a test for presence of a line of pixels is met. When the test is met for a block, that block is identified as pixel-line-present. Pixel-line-present blocks are used to identify blocks that are adjacent. One or more adjacent block(s) may be merged with a pixel-line-present block when one or more rules are found to be satisfied, resulting in a merged block. The merged block is then subject to the above-described test, to verify presence of a line of pixels therein, and when the test is satisfied the merged block is processed normally, e.g. classified as text or non-text.

    Abstract translation: 在所描述的实施例的几个方面中，电子设备和方法使用相机来捕获电子设备外的环境的图像或视频帧，随后识别图像中的区域块。 检查包含区域的每个块，以确定是否满足一行像素的存在测试。 当块的测试被满足时，该块被识别为像素线存在。 像素线存在块用于识别相邻的块。 当发现满足一个或多个规则时，一个或多个相邻块可以与像素线存在块合并，导致合并块。 然后对合并的块进行上述测试，以验证其中的一行像素的存在，并且当满足测试时，合并的块被正常处理，例如， 分类为文本或非文本。

Patent Agency Ranking