专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

US20130185541A1 BITSTREAM BUFFER MANIPULATION WITH A SIMD MERGE INSTRUCTION 有权
标题翻译：使用SIMD MERGE指令进行BITSTREAM BUFFER操作
公开(公告)号：US20130185541A1
公开(公告)日：2013-07-18
申请号：US13788899
申请日：2013-03-07
申请人： YEN-KUANG CHEN , WILLIAM MACY, JR. , MATTHEW HOLLIMAN , ERIC DEBES , MINERVA YEUNG
发明人： YEN-KUANG CHEN , WILLIAM MACY, JR. , MATTHEW HOLLIMAN , ERIC DEBES , MINERVA YEUNG
IPC分类号： G06F9/30
CPC分类号： G06F9/30145 , G06F9/3001 , G06F9/30014 , G06F9/30018 , G06F9/30025 , G06F9/30032 , G06F9/30036 , G06F9/30109 , G06F9/30112 , G06F9/3013 , G06F9/3816 , G06F9/3885 , G06F15/8007 , G06F17/147 , G06F17/15
摘要： Method, apparatus, and program means for performing bitstream buffer manipulation with a SIMD merge instruction. The method of one embodiment comprises determining whether any unprocessed data bits for a partial variable length symbol exist in a first data block is made. A shift merge operation is performed to merge the unprocessed data bits from the first data block with a second data block. A merged data block is formed. A merged variable length symbol comprised of the unprocessed data bits and a plurality of data bits from the second data block is extracted from the merged data block.
摘要翻译：用于使用SIMD合并指令执行比特流缓冲器操作的方法，装置和程序装置。一个实施例的方法包括确定在第一数据块中是否存在用于部分可变长度符号的任何未处理的数据比特。执行移位合并操作以将来自第一数据块的未处理数据位与第二数据块合并。形成合并的数据块。从合并的数据块中提取由未处理的数据位和来自第二数据块的多个数据位组成的合并的可变长度符号。

2. 发明申请

US20050025354A1 Investigation of destroyed assemblies and identification of components thereof 有权
标题翻译：对被毁坏的装配的调查及其部件的识别
公开(公告)号：US20050025354A1
公开(公告)日：2005-02-03
申请号：US10631151
申请日：2003-07-31
申请人： William Macy , Robert Luecking , Ronald Marsden
发明人： William Macy , Robert Luecking , Ronald Marsden
IPC分类号： B23P19/00 , G06K9/00 , G06T7/00
CPC分类号： G06T7/0004 , B33Y50/00 , B33Y80/00
摘要： A method is disclosed for providing a means for identifying recovered component parts of a destroyed assembly quickly and relatively easily using digital or electronic scanning techniques and comparison to virtual components that are presumed to have constituted the original assembly. The method also provides a means for digitally rigging the component parts in three-dimensional virtual space, thereby minimizing and, in some situations, possibly eliminating any need to physical rig the component parts.
摘要翻译：公开了一种用于提供用于使用数字或电子扫描技术快速且相对容易地识别被破坏的组件的恢复的部件的装置的方法，以及与被假定构成原始组件的虚拟部件的比较。该方法还提供了一种在三维虚拟空间中数字地装配组件的方法，从而最小化并且在某些情况下可能消除对物理钻机组件的任何需求。

3. 发明授权

US5822459A Method for processing wavelet bands 失效
标题翻译：处理小波段的方法
公开(公告)号：US5822459A
公开(公告)日：1998-10-13
申请号：US535550
申请日：1995-09-28
申请人： William Macy
发明人： William Macy
IPC分类号： G06F17/14 , G06K9/36
CPC分类号： G06F17/148
摘要： A method and computer system for separating an image into a sequence of wavelet bands, which are equivalent in value to the conventional sequence of wavelet bands, without rearranging the pixel data elements representing the image. The sequence of bands represent different values in reproducing the image. The method of processing the image includes the first step of storing the pixel data elements as first and second packed data sequences in a memory with the pixel data elements stored in an order. A set of intermediate results are generated by performing arithmetic operations between the corresponding pixel data elements of the first and second packed data sequences. The set of intermediate results are then used to generate the sequence of bands by performing arithmetic operations between the intermediate results.
摘要翻译：一种用于将图像分离成小波段的序列的方法和计算机系统，其与传统的小波带序列相当，而不重新排列表示图像的像素数据元素。带的序列在再现图像中代表不同的值。处理图像的方法包括将像素数据元素作为第一和第二打包数据序列存储在具有以顺序存储的像素数据元素的存储器中的第一步骤。通过在第一和第二打包数据序列的相应像素数据元素之间执行算术运算来生成一组中间结果。然后通过在中间结果之间进行算术运算，将该组中间结果用于产生频带序列。

4. 发明授权

US08463820B2 System and method for memory bandwidth friendly sorting on multi-core architectures 有权
标题翻译：多核架构内存带宽友好排序的系统和方法
公开(公告)号：US08463820B2
公开(公告)日：2013-06-11
申请号：US12454883
申请日：2009-05-26
申请人： Jatin Chhugani , Sanjeev Kumar , Anthony-Trung D. Nguyen , Yen-Kuang Chen , Victor W. Lee , William Macy
发明人： Jatin Chhugani , Sanjeev Kumar , Anthony-Trung D. Nguyen , Yen-Kuang Chen , Victor W. Lee , William Macy
IPC分类号： G06F7/00 , G06F17/30 , G06F13/00 , G06F13/28 , G06F12/00
CPC分类号： G06F7/36 , G06F12/0802
摘要： In some embodiments, the invention involves utilizing a tree merge sort in a platform to minimize cache reads/writes when sorting large amounts of data. An embodiment uses blocks of pre-sorted data residing in “leaf nodes” residing in memory storage. A pre-sorted block of data from each leaf node is read from memory and stored in faster cache memory. A tree merge sort is performed on the nodes that are cache resident until a block of data migrates to a root node. Sorted blocks reaching the root node are written to memory storage in an output list until all pre-sorted data blocks have been moved to cache and merged upward to the root. The completed output list in memory storage is a list of the fully sorted data. Other embodiments are described and claimed.
摘要翻译：在一些实施例中，本发明涉及在平台中利用树合并排序以在排序大量数据时最小化高速缓存读/写。实施例使用驻留在存储器存储器中的“叶节点”中驻留的预排序数据块。来自每个叶节点的预先排序的数据块从存储器读取并存储在更快的高速缓冲存储器中。在缓存驻留的节点上执行树合并排序，直到数据块迁移到根节点。到达根节点的排序块被写入到输出列表中的存储器存储器中，直到所有预排序的数据块已被移动到高速缓存并向上合并到根。内存存储器中完成的输出列表是完整排序数据的列表。描述和要求保护其他实施例。

5. 发明授权

US06646681B1 Method for reducing row noise from images 失效
标题翻译：降低图像行噪声的方法
公开(公告)号：US06646681B1
公开(公告)日：2003-11-11
申请号：US09291700
申请日：1999-04-14
申请人： William Macy , Ajaya Durg
发明人： William Macy , Ajaya Durg
IPC分类号： H04N5217
CPC分类号： H04N5/3658
摘要： A method for reducing row noise from a complementary metal oxide semiconductor (CMOS) image sensor is disclosed. The method includes determining a set of row sums for a set of pixel rows in the image sensor and a set of corresponding contributing pixel counts. Then, determining a set of row offset corrections. Finally, adjusting the set of pixel rows by the set of row offset corrections.
摘要翻译：公开了一种用于减少互补金属氧化物半导体（CMOS）图像传感器的行噪声的方法。该方法包括确定图像传感器中的一组像素行的一组行和和一组相应的贡献像素计数。然后，确定一组行偏移校正。最后，通过一组行偏移校正来调整像素行的集合。

6. 发明授权

US06538691B1 Software correction of image distortion in digital cameras 有权
标题翻译：数码相机图像失真的软件校正
公开(公告)号：US06538691B1
公开(公告)日：2003-03-25
申请号：US09235136
申请日：1999-01-21
申请人： William Macy , Oleg Rashkovskiy
发明人： William Macy , Oleg Rashkovskiy
IPC分类号： H04N5235
CPC分类号： H04N17/002 , H04N3/2335 , H04N5/3572
摘要： A digital image processing method and apparatus corrects geometrical distortions in digital images. The method includes the calibration of the camera by choosing a test pattern and manually adjusting the distortion correction factor of the camera until the image appears to be undistorted. A real image is then captured by the camera grid. The camera grid is then distorted by the same distortion factor such that the image is corrected, but the image contour is deformed. Based on the distortion correction factor, a respective location in the camera grid is determined for each pixel in a display window. An intensity value to each pixel in the display window is assigned based on one or more pixels in the camera grid indicated by the location. The points for mapping camera values to the display window need to be determined only once for a given image size. All subsequent distorted images having the same size are corrected using interpolation and pre-compiled points.
摘要翻译：数字图像处理方法和装置校正数字图像中的几何失真。该方法包括通过选择测试图案并手动调整相机的失真校正系数来校准相机，直到图像看起来不失真。然后，相机网格捕获真实图像。相机网格然后被相同的失真因子扭曲，使得图像被校正，但图像轮廓变形。基于失真校正因子，在显示窗口中为每个像素确定摄像机网格中的相应位置。基于由位置指示的摄像机网格中的一个或多个像素来分配显示窗口中每个像素的强度值。将相机值映射到显示窗口的点只需要为给定的图像尺寸确定一次。使用插值和预编译点来校正具有相同尺寸的所有随后的失真图像。

7. 发明申请

US20110066806A1 System and method for memory bandwidth friendly sorting on multi-core architectures 有权
标题翻译：多核架构内存带宽友好排序的系统和方法
公开(公告)号：US20110066806A1
公开(公告)日：2011-03-17
申请号：US12454883
申请日：2009-05-26
申请人： Jatin Chhugani , Sanjeev Kumar , Anthony-Trung D. Nguyen , Yen-Kuang Chen , Victor W. Lee , William Macy
发明人： Jatin Chhugani , Sanjeev Kumar , Anthony-Trung D. Nguyen , Yen-Kuang Chen , Victor W. Lee , William Macy
IPC分类号： G06F12/08
CPC分类号： G06F7/36 , G06F12/0802
摘要： In some embodiments, the invention involves utilizing a tree merge sort in a platform to minimize cache reads/writes when sorting large amounts of data. An embodiment uses blocks of pre-sorted data residing in “leaf nodes” residing in memory storage. A pre-sorted block of data from each leaf node is read from memory and stored in faster cache memory. A tree merge sort is performed on the nodes that are cache resident until a block of data migrates to a root node. Sorted blocks reaching the root node are written to memory storage in an output list until all pre-sorted data blocks have been moved to cache and merged upward to the root. The completed output list in memory storage is a list of the fully sorted data. Other embodiments are described and claimed.
摘要翻译：在一些实施例中，本发明涉及在平台中利用树合并排序以在排序大量数据时最小化高速缓存读/写。实施例使用驻留在存储器存储器中的“叶节点”中驻留的预排序数据块。来自每个叶节点的预先排序的数据块从存储器读取并存储在更快的高速缓冲存储器中。在缓存驻留的节点上执行树合并排序，直到数据块迁移到根节点。到达根节点的排序块被写入到输出列表中的存储器存储器中，直到所有预排序的数据块已被移动到高速缓存并向上合并到根。内存存储器中完成的输出列表是完整排序数据的列表。描述和要求保护其他实施例。

8. 发明申请

US20050108312A1 Bitstream buffer manipulation with a SIMD merge instruction 有权
标题翻译：使用SIMD合并指令进行位流缓冲操作
公开(公告)号：US20050108312A1
公开(公告)日：2005-05-19
申请号：US10612542
申请日：2003-07-01
申请人： Yen-Kuang Chen , William Macy , Matthew Holliman , Eric Debes , Minerva Young
发明人： Yen-Kuang Chen , William Macy , Matthew Holliman , Eric Debes , Minerva Young
IPC分类号： G06F9/30 , G06F9/302 , G06F9/308 , G06F9/315 , G06F9/38 , G06F17/14 , G06F7/38
CPC分类号： G06F9/30145 , G06F9/3001 , G06F9/30014 , G06F9/30018 , G06F9/30025 , G06F9/30032 , G06F9/30036 , G06F9/30109 , G06F9/30112 , G06F9/3013 , G06F9/3816 , G06F9/3885 , G06F15/8007 , G06F17/147 , G06F17/15
摘要： Method, apparatus, and program means for performing bitstream buffer manipulation with a SIMD merge instruction. The method of one embodiment comprises determining whether any unprocessed data bits for a partial variable length symbol exist in a first data block is made. A shift merge operation is performed to merge the unprocessed data bits from the first data block with a second data block. A merged data block is formed. A merged variable length symbol comprised of the unprocessed data bits and a plurality of data bits from the second data block is extracted from the merged data block.
摘要翻译：用于使用SIMD合并指令执行比特流缓冲器操作的方法，装置和程序装置。一个实施例的方法包括确定在第一数据块中是否存在用于部分可变长度符号的任何未处理的数据比特。执行移位合并操作以将来自第一数据块的未处理数据位与第二数据块合并。形成合并的数据块。从合并的数据块中提取由未处理的数据位和来自第二数据块的多个数据位组成的合并的可变长度符号。

9. 发明授权

US06181376B2 Method of determining missing color values for pixels in a color filter array 失效
标题翻译：确定滤色器阵列中像素缺失颜色值的方法
公开(公告)号：US06181376B2
公开(公告)日：2001-01-30
申请号：US08950166
申请日：1997-10-14
申请人： Oleg Rashkovskiy , William Macy
发明人： Oleg Rashkovskiy , William Macy
IPC分类号： H04N314
CPC分类号： G06T3/4015 , H04N9/045
摘要： Missing color values for pixels in a color filter array created by a digital camera and having a Bayer pattern are generated by determining values for all missing green color values for pixels in the color filter array by interpolation of known green color values adjacent along diagonal lines and determining values for all missing red and blue color values for the pixels from the sum of an interpolation term and a luminance correction term based on differences between green color values of adjacent pixels. The missing green color values are determined by computing temporary values for green which lie at the corners of pixels having known green color values using cubic B-spline filters oriented along diagonally adjacent green color pixels, determining final values for green at the centers of pixels having known red and blue values by using cubic B-spline filters oriented along diagonally adjacent red and blue color pixels, and by sampling the temporary green values.
摘要翻译：通过数字相机创建并具有拜耳图案的滤色器阵列中的像素的缺失颜色值是通过对对角线相邻的已知绿色颜色值的插值来确定滤色器阵列中的像素的所有缺少的绿色颜色值的值而生成的，基于相邻像素的绿色颜色值之间的差异，根据插值项和亮度校正项的和来确定像素的所有缺失的红色和蓝色颜色值的值。丢失的绿色颜色值通过使用沿对角线相邻的绿色像素定向的三次B样条滤波器来计算位于具有已知绿色颜色值的像素的角落处的绿色临时值，确定在具有通过使用沿对角线相邻的红色和蓝色像素定向的三次B样条滤波器，并通过对临时绿色值进行采样来获得已知的红色和蓝色值。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式