公开(公告)号：US20080180749A1

公开(公告)日：2008-07-31

申请号：US11627305

申请日：2007-01-25

申请人： Stephen POLLARD ,  Andrew HUNTER

发明人： Stephen POLLARD ,  Andrew HUNTER

IPC分类号： H04N1/40

CPC分类号： H04N1/3871 ,  G06T5/50 ,  H04N1/4092 ,  H04N5/235 ,  H04N5/2355

摘要： An image processing system includes: an image capture device operable to capture at least first and second images of an object at different exposures, a storage for at least part of the at least first and second images, a filter for splitting the at least first and second images into respective first and second spatial frequency components, the first frequency component being of lower frequency than the second, a selector for selecting local image intensity data from the second spatial frequency components of a region of the at least first and second images for combination with the first frequency component from the corresponding region of the first image, and a combiner for combining the selected local image data from the second spatial frequency components of the at least first and second images with the first spatial frequency component of the first image to form a final image.

摘要翻译： 一种图像处理系统包括：图像捕获装置，其可操作以在不同的曝光下捕获对象的至少第一和第二图像，至少部分至少第一和第二图像的存储;用于将至少第一和第二图像分割的滤波器; 第二图像分成相应的第一和第二空间频率分量，第一频率分量的频率低于第二频率分量;选择器，用于从用于组合的至少第一和第二图像的区域的第二空间频率分量中选择局部图像强度数据 来自第一图像的相应区域的第一频率分量和用于将来自至少第一和第二图像的第二空间频率分量的所选局部图像数据与第一图像的第一空间频率分量组合的组合器以形成 最后的形象。

公开(公告)号：US20080100878A1

公开(公告)日：2008-05-01

申请号：US11930981

申请日：2007-10-31

申请人： Andrew HUNTER ,  Stephen POLLARD

发明人： Andrew HUNTER ,  Stephen POLLARD

IPC分类号： H04N1/387

CPC分类号： H04N1/484 ,  H04N1/3871

摘要： A method of generating an image of a portion of an object, the method comprising generating a plurality of images of the portion of the object, each image being represented by a respective set of image data, applying a comparative data transform to at least a portion of each set of image data such that the transformed comparative image data isolates variations in the illumination between the plurality of images, determining a plurality of normalisation transforms from the transformed comparative image data such that when applied to a respective one of the plurality of image data sets minimises the variation in illumination applying each determined normalisation transform to the respective set of image data and transforming the normalised sets of image data to a single image of the portion of the object.

摘要翻译： 一种产生对象的一部分的图像的方法，所述方法包括生成所述对象的所述部分的多个图像，每个图像由相应的图像数据集合表示，将比较数据变换应用于至少一部分 使得变换的比较图像数据隔离多个图像之间的照明的变化，从转换的比较图像数据确定多个归一化变换，使得当应用于多个图像数据中的相应一个图像数据时 将照射应用每个确定的归一化变换的照明变化最小化到相应的图像数据集合，并将归一化的图像数据集合变换成对象的该部分的单个图像。

公开(公告)号：US20120212324A1

公开(公告)日：2012-08-23

申请号：US13032479

申请日：2011-02-22

申请人： Stephen POLLARD ,  Steven J. Simske

发明人： Stephen POLLARD ,  Steven J. Simske

IPC分类号： G06K9/66 ,  G05B19/00

CPC分类号： G06K9/00161 ,  G06K9/00577 ,  G06K9/48 ,  G06K2209/01

摘要： A method for model-based signature profile extraction includes capturing an image of an authentic glyph. An outline model is fit to the image of the authentic glyph, and an authentic signature profile is extracted based on the outline model. A signature profile extracted from an image of another glyph may be compared to the to the authentic signature profile so as to forensically verify authenticity of the other glyph The system for model-based signature profile extraction includes a controller, a capture unit, an outline unit, a profiling unit, and a forensic verification unit. A computer readable medium containing executable instructions is also described.

摘要翻译： 用于基于模型的签名简档提取的方法包括捕获真实字形的图像。 轮廓模型适合于真实字形的图像，并且基于轮廓模型提取真实的签名简档。 从另一个字形的图像提取的签名简档可以与真实的签名简档进行比较，以便对其他字形的真实性进行验证。用于基于模型的签名简档提取的系统包括控制器，捕获单元，轮廓单元 ，分析单位和法医验证单位。 还描述了包含可执行指令的计算机可读介质。