会员体验
专利管家(专利管理)
工作空间(专利管理)
风险监控(情报监控)
数据分析(专利分析)
侵权分析(诉讼无效)
联系我们
交流群
官方交流:
QQ群: 891211   
微信请扫码    >>>
现在联系顾问~
热词
    • 5. 发明授权
    • Method and system for modeling handwriting using polynomials as a
function of time
    • 使用多项式作为时间的函数建模笔迹的方法和系统
    • US5764797A
    • 1998-06-09
    • US428412
    • 1995-04-25
    • James L. Adcock
    • James L. Adcock
    • G06K9/22G06K9/52G06K9/00G06K9/36G06K9/46G06K9/48
    • G06K9/52G06K9/00409G06K9/00429G06K2209/01
    • A method and system for modeling handwriting using polynomials as a function of time. In a preferred embodiment, a user inputs handwriting by writing with an electronic pen on an electronic tablet. The tablet samples the location of the tip of the pen on the tablet from pen down until pen up at periodic intervals as each stroke in a character is written and inputs the (x,y) coordinates for each stroke in the character to a normalizer. The normalizer normalizes the (x,y) coordinates and inputs the normalized (x,y) coordinates to a featurizer. The featurizer generates a set of features that represent each stroke in the character. In the preferred embodiment, the features are coefficients of a Chebyshev. After the Chebyshev coefficients have been generated for each stroke in the character, the featurizer normalizes the set of Chebyshev coefficients. During training, the normalized Chebyshev coefficients are input to a prototype generator. After receiving all of the input samples for a given character and the corresponding set of Chebyshev coefficients for each sample, the prototype generator generates one or more prototypes of the given character and stores the prototype(s) in a database. During recognition, the normalized Chebyshev coefficients are input to a classifier. After receiving the set of Chebyshev coefficients for a character to be recognized, the classifier compares the set of Chebyshev coefficients to the prototypes stored in the database to find an acceptable match. If an acceptable match is found, the character corresponding to the matching prototype is output.
    • 一种用于使用多项式作为时间的函数建模手写的方法和系统。 在优选实施例中,用户通过用电子笔在电子平板电脑上书写输入笔迹。 平板电脑将笔尖上的位置从笔直下笔直到笔直到笔直到笔写为一个字符中的每个笔划,并将字符中的每个笔画的(x,y)坐标输入到归一化器。 归一化器对(x,y)坐标进行归一化,并将归一化(x,y)坐标输入到特征化器。 特征化器生成一组表示字符中每个笔画的特征。 在优选实施例中,特征是切比雪夫的系数。 在为字符中的每个笔划生成切比雪夫系数之后,特征化器将切比雪夫系数集合归一化。 在训练期间,归一化的切比雪夫系数被输入到原型发生器。 在接收到给定字符的所有输入样本和每个样本的相应的切比雪夫系数集合之后,原型生成器生成给定字符的一个或多个原型并将原型存储在数据库中。 在识别期间,将归一化的切比雪夫系数输入到分类器。 在接收到要识别的字符的切比雪夫系数集合之后,分类器将切比雪夫系数集合与存储在数据库中的原型进行比较以找到可接受的匹配。 如果发现可接受的匹配,则输出与匹配原型对应的字符。
    • 10. 发明授权
    • Noise corrected pole and zero analyzer
    • 噪声校正极和零分析仪
    • US4654809A
    • 1987-03-31
    • US644307
    • 1984-08-23
    • James L. Adcock
    • James L. Adcock
    • G06F17/10G06F15/20G06F11/00
    • G06F17/10
    • A pole and zero analyzer determines the poles and zeroes of a device by applying a stimulus signal to the device, detecting the stimulus and response signals and computing the auto- and cross spectra. An estimated transfer function comprising a rational fraction of two Chebyshev polynomials is least squares fit to a measured transfer function obtained from the auto- and cross-spectra and the poles and zeroes are obtained as the roots of the estimated transfer function. Inaccuracies due to noise and computational error are minimized by the use of the Chebyshev polynomials.
    • 极和零分析仪通过向设备施加刺激信号,检测刺激和响应信号以及计算自动和交叉光谱来确定设备的极点和零点。 包含两个切比雪夫多项式的有理分数的估计传递函数是与从自动和交叉频谱获得的测量的传递函数拟合的最小二乘法,并且获得作为估计传递函数的根的极点和零点。 由于噪声和计算误差引起的不准确性通过使用切比雪夫多项式来最小化。