公开(公告)号：US20070286517A1

公开(公告)日：2007-12-13

申请号：US11639964

申请日：2006-12-15

申请人： Joon-Ki Paik ,  Jeong-Ho Shin ,  Maik Vivek ,  Yoon-Jong Yoo

发明人： Joon-Ki Paik ,  Jeong-Ho Shin ,  Maik Vivek ,  Yoon-Jong Yoo

IPC分类号： G06K9/46

CPC分类号： H04N5/23212 ,  G06T5/003 ,  H04N5/23232

摘要： Methods and apparatus are disclosed for digital image restoration. The method including the steps of receiving an external deteriorated image, and generating N restoration images by using a preset PSF and an image restoration filter without computing a PSF of the input deteriorated image, dividing the restoration images into window parts of a predetermined window size, computing absolute values of differences among peripheral pixels within the divided window parts, selecting only a high frequency part, and analyzing focus deterioration; determining weights of the N restoration images, selecting window parts having high weights, and integrating the selected window parts into one image and comparing a degree of restoration of the integrated image with a preset degree of restoration, and outputting the integrated image as a resultant image when the degree of restoration of the integrated image is higher than the preset degree of restoration.

摘要翻译： 公开了用于数字图像恢复的方法和装置。 该方法包括以下步骤：接收外部劣化图像，并且在不计算输入劣化图像的PSF的情况下通过使用预设PSF和图像恢复滤波器生成N个恢复图像，将恢复图像划分为预定窗口大小的窗口部分， 计算分割窗口部分之间的周边像素之间的差异的绝对值，仅选择高频部分，并分析焦点劣化; 确定所述N个恢复图像的权重，选择具有高权重的窗口部分，以及将所选择的窗口部分整合为一个图像，并将所述集成图像的恢复程度与预设恢复度进行比较，并将所述集成图像作为合成图像 当整合图像的恢复程度高于预设的恢复程度时。

公开(公告)号：US5307167A

公开(公告)日：1994-04-26

申请号：US851498

申请日：1992-03-16

申请人： Sung-Wook Park ,  Joon-Ki Paik

发明人： Sung-Wook Park ,  Joon-Ki Paik

IPC分类号： G06F3/153 ,  G06T3/40 ,  G09G5/36 ,  H04N1/393 ,  H04N5/228 ,  H04N3/223 ,  H04N5/14

CPC分类号： G06T3/4007

摘要： The present invention relates to a digital zoom system utilizing image buffers and employing an approximated bilinear interpolation method. The zoom system includes an analog-to-digital (A/D) converter for converting an analog image signal to a digital image signal, an image buffer for storing the digital image signal generated for said A/D converter, a horizontal operation circuit for interpolating the digital image signal taken along with a horizontal direction, one or more horizontal image buffers for storing the digital image signal for 1-horizontal lines magnified in a horizontal direction, a vertical operation circuit for interpolating said horizontally magnified image signal in a vertical direction, a timing controller for controlling the image buffer, said horizontal and vertical operation circuits, and a digital-to-analog (D/A) converter for converting the digital image signal resultedly magnified in the horizontal and vertical direction back to analog image signal.

摘要翻译： 本发明涉及利用图像缓冲器并采用近似双线性插值方法的数字变焦系统。 变焦系统包括用于将模拟图像信号转换为数字图像信号的模数（A / D）转换器，用于存储为所述A / D转换器生成的数字图像信号的图像缓冲器，用于 内插水平方向取得的数字图像信号，一个或多个水平图像缓冲器，用于存储水平方向放大的1水平线的数字图像信号;垂直运算电路，用于沿垂直方向内插所述水平放大图像信号 ，用于控制图像缓冲器的定时控制器，所述水平和垂直运算电路，以及数字 - 模拟（D / A）转换器，用于将在水平和垂直方向上被成倍放大的数字图像信号转换回模拟图像信号。

公开(公告)号：US08526722B2

公开(公告)日：2013-09-03

申请号：US13088070

申请日：2011-04-15

申请人： Joon-Ki Paik ,  Eun-Sung Lee

发明人： Joon-Ki Paik ,  Eun-Sung Lee

IPC分类号： G06K9/00

CPC分类号： H04N5/2356 ,  H04N5/23212 ,  H04N9/045

摘要： Provided are an apparatus and method for enhancing image quality of an input image captured by an image capturing device having an aperture with a plurality of openings employing different color filters. The image quality enhancing apparatus includes an ROI (region of interest) classifying unit classifying ROIs according to distances from the image capturing device of a captured input image by using a cluster-based image segmentation algorithm, a color channel registering unit determining motion vectors of color channels corresponding to the respective ROIs by analyzing a high frequency component of image data included in the ROIs, and moving the color channels based on the determined motion vectors to obtain a registered image of the respective ROIs, and an image fusing unit fusing the registered images of the respective ROIs to obtain a fused image.

摘要翻译： 提供了一种用于增强由具有带有多个采用不同滤色器的开口的孔径的图像捕获装置拍摄的输入图像的图像质量的装置和方法。 图像质量提高装置包括ROI（感兴趣区域）分类单元，通过使用基于聚类的图像分割算法，根据捕获的输入图像的图像捕获装置的距离对ROI进行分类，颜色通道登记单元确定颜色的运动矢量 通过分析包括在ROI中的图像数据的高频分量，并且基于所确定的运动矢量来移动颜色通道以获得各ROI的登记图像，对应于相应ROI的通道，以及融合注册图像的图像定影单元 以获得融合图像。

公开(公告)号：US07885462B2

公开(公告)日：2011-02-08

申请号：US11562535

申请日：2006-11-22

申请人： Joon-Ki Paik

发明人： Joon-Ki Paik

IPC分类号： G06K9/00 ,  G06K9/40 ,  G06K9/38 ,  H04N1/40

CPC分类号： G06T5/40 ,  G06T5/009

摘要： An image processing method and system using gain-controllable clipped histogram equalization, in which the image processing method includes the steps of as obtaining a brightness histogram, computing a mean brightness of an image signal, determining a clipping rate based on the mean brightness, determining a clipping threshold based on the clipping rate, obtaining a clipped brightness histogram by clipping frequencies exceeding the clipping threshold in the brightness histogram, obtaining a corrected brightness histogram by correcting the clipped brightness as histogram using the clipping rate as a total gain, obtaining a cumulative histogram from the corrected brightness histogram, and correcting an input image using the cumulative histogram as a transformation function. Accordingly, the clipping rate is adaptively controlled so that image contrast is enhanced.

摘要翻译： 一种使用增益可控削减直方图均衡的图像处理方法和系统，其中图像处理方法包括以下步骤：获得亮度直方图，计算图像信号的平均亮度，基于平均亮度确定限幅率，确定 基于剪切速率的剪切阈值，通过限制亮度直方图中的限幅阈值的频率来获得剪切亮度直方图，通过使用限幅率校正剪辑亮度作为总增益获得校正的亮度直方图，获得累积 并且使用累积直方图校正输入图像作为变换函数。 因此，限幅率被自适应地控制，从而增强了图像对比度。

公开(公告)号：US20090231447A1

公开(公告)日：2009-09-17

申请号：US12402605

申请日：2009-03-12

申请人： Joon-Ki Paik ,  Kyoung-Ju Park ,  In-Ho Paek

发明人： Joon-Ki Paik ,  Kyoung-Ju Park ,  In-Ho Paek

IPC分类号： H04N5/222

CPC分类号： G01S3/7864 ,  G06T3/4038 ,  H04N5/23238

摘要： Provided are an apparatus and method for generating panorama images and an apparatus and method for tracking an object using the same. The apparatus for generating panorama images estimates regional motion vectors with respect to respective lower regions set in an image frame, determines a frame motion vector by using the regional motion vectors based on a codebook storing sets of normalized regional motion vectors corresponding to a camera's motion and then accumulates motion mediating variables computed with respect to respective continuous image frames to determine an overlapping region, and matches the overlapping regions to generate a panorama image. The apparatus for tracking an object sets the panorama image generated by the apparatus for generating panorama images as a background image, and determines an object region by averaging pixel value difference values between pixels that constitute a captured input image and corresponding pixels in the panorama image. Regional motion vectors are estimated by using the coordinate of a pixel selected in lower regions, and thus a computing process can be simplified and a process speed can be increased. In addition, effects of wrongly estimated regional motion vectors can be minimized by using a codebook. Furthermore, noises are minimized and an object can be accurately extracted by using averaged pixel value difference values.

摘要翻译： 提供了一种用于产生全景图像的装置和方法，以及用于跟踪使用该图像的对象的装置和方法。 用于生成全景图像的装置估计相对于在图像帧中设置的相应较低区域的区域运动矢量，通过使用区域运动矢量来确定帧运动矢量，所述区域运动矢量基于存储与相机运动对应的归一化区域运动矢量集合的码本， 然后累积关于相应连续图像帧计算的运动中介变量以确定重叠区域，并匹配重叠区域以生成全景图像。 用于跟踪对象的装置将由生成全景图像的装置生成的全景图像设置为背景图像，并且通过对构成捕获的输入图像的像素与全景图像中的对应像素之间的像素值差值进行平均来确定对象区域。 通过使用在较低区域中选择的像素的坐标来估计区域运动矢量，从而可以简化计算过程并且可以提高处理速度。 此外，通过使用码本可以最大限度地减少错误估计的区域运动矢量的影响。 此外，噪声被最小化，并且可以通过使用平均像素值差值来精确地提取对象。

公开(公告)号：US06501460B1

公开(公告)日：2002-12-31

申请号：US09991700

申请日：2001-11-26

申请人： Joon-Ki Paik ,  Dong-Hoon Kang

发明人： Joon-Ki Paik ,  Dong-Hoon Kang

IPC分类号： G09G508

CPC分类号： G06F3/0317

摘要： A light-receiving unit which is simplified by dispensing with a case or a lens and thus improved in manufacturing efficiency and lowered in manufacturing cost is disclosed. The light-receiving unit comprises: a lens for converging light, which is emitted from a light source and then reflected from a contacting object, in which the light carries an image formed on the contacting object; a printed circuit board having a light hole for allowing the light converged by the lens to be passed therethrough; and a molding package made by molding lead frames electrically connected to the printed circuit board and an optical sensor bonded to the lead frames with transparent resin, which is adapted to detect the image carried in the converged light.

摘要翻译： 公开了一种通过分配壳体或透镜而被简化并因此提高了制造效率并降低了制造成本的光接收单元。 光接收单元包括：用于会聚光的透镜，其从光源发射然后从接触物体反射，其中光携带形成在接触物体上的图像; 印刷电路板，具有用于使透镜会聚的光通过的光孔; 以及通过模制电连接到印刷电路板的引线框和通过透明树脂结合到引线框的光学传感器制成的成型封装，其适于检测在会聚光中承载的图像。

公开(公告)号：US08644604B2

公开(公告)日：2014-02-04

申请号：US13356842

申请日：2012-01-24

申请人： Joon-Ki Paik ,  Jin-Hee Lee

发明人： Joon-Ki Paik ,  Jin-Hee Lee

IPC分类号： G06K9/34

CPC分类号： H04N9/045 ,  G06T5/50 ,  G06T7/33 ,  G06T2207/10024 ,  G06T2207/20221 ,  H04N5/23212 ,  H04N5/23254

摘要： Provided are an apparatus and method for aligning color channels on the basis of depth information on an image taken by an imaging device equipped with a multiple color-filter aperture (MCA). An image divider generates a binary image including edge information on an original image, and generates a label map including a plurality of regions of interest (ROIs) respectively corresponding to a plurality of objects included in the original image. A shifting vector estimator estimates color shifting vectors (CSVs) indicating shift directions and distances of color channels of the original image according to the respective ROIs. A channel matching unit obtains a plurality of matched images corresponding to the respective ROIs included in the label map by shifting the color channels according to the respective CSVs, and generates a multi-focus image by combining the matched images.

摘要翻译： 提供了一种用于基于由配备有多个彩色滤光片孔径（MCA）的成像装置拍摄的图像上的深度信息来对准颜色通道的装置和方法。 图像分割器生成包括原始图像上的边缘信息的二进制图像，并且生成包括分别对应于包括在原始图像中的多个对象的多个感兴趣区域（ROI）的标签映射。 移动矢量估计器根据各个ROI估计指示原始图像的颜色通道的移动方向和距离的颜色移动向量（CSV）。 信道匹配单元通过根据各个CSV移位颜色信道来获得与包括在标签映射中的各个ROI相对应的多个匹配图像，并且通过组合匹配图像来生成多焦点图像。

公开(公告)号：US06541762B2

公开(公告)日：2003-04-01

申请号：US09984283

申请日：2001-10-29

申请人： Dong-Hoon Kang ,  Joon-Ki Paik

发明人： Dong-Hoon Kang ,  Joon-Ki Paik

IPC分类号： H01J4014

CPC分类号： G06F3/0317 ,  H05K1/0274 ,  H05K1/141 ,  H05K3/3447 ,  H05K2201/09072 ,  H05K2201/10303 ,  H05K2201/10477 ,  H05K2201/10871

摘要： A sub chip on board for an optical mouse is disclosed. The chip on board has a sub PCB having both a plurality of input/output pads and a plurality of pin holes. A sensor die has an optical sensor wire-bonded to said input/output pads for sensing received light, and is attached to a bottom surface of the sub PCB. A transparent resin covers the sensor die at the bottom surface of the sub PCB. A cap is attached to the bottom surface of the sub PCB such that the cap covers the transparent resin, and has a hole for guiding the received light to said optical sensor. A main PCB has both a hole for guiding the received light to the optical sensor and a plurality of pin holes corresponding to the pin holes of the sub PCB. A plurality of pins are commonly inserted into the pin holes of both the main PCB and the sub PCB.

摘要翻译： 公开了一种用于光电鼠标的子芯片。 板上的芯片具有具有多个输入/输出焊盘和多个销孔的子PCB。 传感器管芯具有导线接合到所述输入/输出焊盘的光学传感器，用于感测接收到的光，并且附接到子PCB的底表面。 透明树脂覆盖子PCB底面的传感器芯片。 盖子附接到副PCB的底表面，使得帽覆盖透明树脂，并且具有用于将接收的光引导到所述光学传感器的孔。 主PCB具有用于将接收的光引导到光学传感器的孔和与子PCB的针孔对应的多个针孔。 多个引脚通常插入主PCB和副PCB的引脚孔中。

公开(公告)号：US08023003B2

公开(公告)日：2011-09-20

申请号：US12402605

申请日：2009-03-12

申请人： Joon-Ki Paik ,  Kyoung-Ju Park ,  In-Ho Paek

发明人： Joon-Ki Paik ,  Kyoung-Ju Park ,  In-Ho Paek

IPC分类号： H04N5/228 ,  H04N7/00 ,  G06K9/32

CPC分类号： G01S3/7864 ,  G06T3/4038 ,  H04N5/23238

摘要： Provided are an apparatus and method for generating panorama images and an apparatus and method for tracking an object using the same. The apparatus for generating panorama images estimates regional motion vectors with respect to respective lower regions set in an image frame, determines a frame motion vector by using the regional motion vectors based on a codebook storing sets of normalized regional motion vectors corresponding to a camera's motion and then accumulates motion mediating variables computed with respect to respective continuous image frames to determine an overlapping region, and matches the overlapping regions to generate a panorama image.

摘要翻译： 提供了一种用于产生全景图像的装置和方法，以及用于跟踪使用该图像的对象的装置和方法。 用于生成全景图像的装置估计相对于在图像帧中设置的相应较低区域的区域运动矢量，通过使用区域运动矢量来确定帧运动矢量，所述区域运动矢量基于存储与相机运动对应的归一化区域运动矢量集合的码本， 然后累积关于相应连续图像帧计算的运动中介变量以确定重叠区域，并匹配重叠区域以生成全景图像。

公开(公告)号：US07881551B2

公开(公告)日：2011-02-01

申请号：US11639964

申请日：2006-12-15

申请人： Joon-Ki Paik ,  Jeong-Ho Shin ,  Maik Vivek ,  Yoon-Jong Yoo

发明人： Joon-Ki Paik ,  Jeong-Ho Shin ,  Maik Vivek ,  Yoon-Jong Yoo

IPC分类号： G06K9/40

CPC分类号： H04N5/23212 ,  G06T5/003 ,  H04N5/23232

摘要： Methods and apparatus are disclosed for digital image restoration. The method including the steps of receiving an external deteriorated image, and generating N restoration images by using a preset PSF and an image restoration filter without computing a PSF of the input deteriorated image, dividing the restoration images into window parts of a predetermined window size, computing absolute values of differences among peripheral pixels within the divided window parts, selecting only a high frequency part, and analyzing focus deterioration; determining weights of the N restoration images, selecting window parts having high weights, and integrating the selected window parts into one image and comparing a degree of restoration of the integrated image with a preset degree of restoration, and outputting the integrated image as a resultant image when the degree of restoration of the integrated image is higher than the preset degree of restoration.

摘要翻译： 公开了用于数字图像恢复的方法和装置。 该方法包括以下步骤：接收外部劣化图像，并且在不计算输入劣化图像的PSF的情况下通过使用预设PSF和图像恢复滤波器生成N个恢复图像，将恢复图像划分为预定窗口大小的窗口部分， 计算分割窗口部分之间的周边像素之间的差异的绝对值，仅选择高频部分，并分析焦点劣化; 确定所述N个恢复图像的权重，选择具有高权重的窗口部分，以及将所选择的窗口部分整合为一个图像，并将所述集成图像的恢复程度与预设恢复度进行比较，并将所述集成图像作为合成图像 当整合图像的恢复程度高于预设的恢复程度时。