公开(公告)号：US06532468B2

公开(公告)日：2003-03-11

申请号：US09215556

申请日：1998-12-18

Applicant: Yoshihiro Ishida ,  Masahiro Konishi

Inventor： Yoshihiro Ishida ,  Masahiro Konishi

IPC: G06F1700

CPC classification number: G06F17/30988 ,  Y10S707/99942 ,  Y10S707/99945

Abstract: In a binary search, two storage units are prepared so that when the least significant bit of the search address is “0” and “1”, even and odd address banks respectively are used. The search object data is classified according to data belonging to the odd and even addresses in continuous addresses and allocated to these two storage units. Further, a search tree of the search address is constructed so that two data of object for a next search are stored in different storage units. Upon the binary search, addresses for the two storage units are set according to this search tree. Therefore, simultaneous readout of data is enabled, so that readout and comparison are carried out in parallel. Further, according to multiple division search of the invention, if 2 bits of the least significant bits of the search address are “00”, “01”, “10” and “11”, a search object data is stored in first through fourth banks respectively. A search tree of the search address is constructed so that four data of object for a next search are stored in different banks. Upon a search, an address for each bank is set according to the search tree so as to enable readout of data at the same time, thereby reducing time required for the search.

Abstract translation: 在二进制搜索中，准备两个存储单元，使得当搜索地址的最低有效位为“0”和“1”时，分别使用偶数和奇数地址组。 搜索对象数据根据属于连续地址中的奇数和偶数地址的数据分类并分配给这两个存储单元。 此外，搜索地址的搜索树被构造成使得用于下一次搜索的对象的两个数据被存储在不同的存储单元中。 在二进制搜索中，根据该搜索树设置两个存储单元的地址。 因此，能够同时读出数据，从而并行地执行读出和比较。 此外，根据本发明的多重搜索，如果搜索地址的最低有效位的2比特是“00”，“01”，“10”和“11”，则搜索对象数据被存储在第一到第四 银行。 构造搜索地址的搜索树，使得用于下一个搜索的对象的四个数据被存储在不同的存储体中。 在搜索时，根据搜索树设置每个存储体的地址，以便能够同时读取数据，从而减少搜索所需的时间。

公开(公告)号：US06457001B1

公开(公告)日：2002-09-24

申请号：US09274809

申请日：1999-03-23

Applicant: Yoshihiro Ishida

Inventor： Yoshihiro Ishida

IPC: G06F1730

CPC classification number: G06F17/30985 ,  G06F7/02 ,  Y10S707/955 ,  Y10S707/99933 ,  Y10S707/99936 ,  Y10S707/99943

Abstract: A method and apparatus for carrying out binary search based on a plurality of retrieval conditions. Retrieval object data is classified into higher order bits and lower order bits. A portion of retrieval key data, in correspondence with the higher order bits of the retrieval object data, constitutes first key data and a portion in correspondence with the lower order bits of the retrieval object data constitutes second retrieval key data. A higher order bit comparing circuit 13 is provided for comparing the higher order bits of the retrieval object data with the first retrieval key data and a lower order bit comparing circuit 14 and for comparing the lower order bits of the retrieval object data with the second retrieval key data. Further, there is provided a determining circuit 16 for determining coincidence or non-coincidence between the retrieval key data and the retrieval object data in accordance with a comparison enabling bit EB indicating coincidence conditions, a comparison result of the higher order bit comparing circuit 13 and comparison result of the lower bit comparing circuit 14.

公开(公告)号：US06289136B1

公开(公告)日：2001-09-11

申请号：US08619910

申请日：1996-03-20

Applicant: Takahiro Oshino ,  Yoshihiro Ishida ,  Shinichiro Koga ,  Osamu Morimoto

Inventor： Takahiro Oshino ,  Yoshihiro Ishida ,  Shinichiro Koga ,  Osamu Morimoto

IPC: G06K932

CPC classification number: G06T3/403

Abstract: In magnifying a multivalued image, binary images are generated by using the respective density levels of the multivalued image as thresholds. The contour of each binary image is equivalent to an equidensity curve corresponding to each density gradient of the original multivalued image. Outline vectors of each binary image are extracted and magnified/smoothed. The contour of each binary image represents an equidensity curve corresponding to each density gradient of the magnified multivalued image. In a binary image area corresponding to a black pixel of each binary image, a pixel having a density equal to the threshold used to generate the binary image is written. This processing is sequentially performed for all the binary images from a low-density binary image to a high-density binary image to generate a multivalued image. The objected multivalued image is subjected to density gradient smoothing processing. As a result, a magnified multivalued image is obtained.

Abstract translation: 在放大多值图像时，通过使用多值图像的相应浓度水平作为阈值来生成二进制图像。 每个二进制图像的轮廓等价于对应于原始多值图像的每个密度梯度的等距曲线。 提取并放大/平滑每个二进制图像的轮廓向量。 每个二进制图像的轮廓表示对应于放大的多值图像的每个密度梯度的等距曲线。 在对应于每个二进制图像的黑色像素的二进制图像区域中，写入具有等于用于生成二值图像的阈值的浓度的像素。 对于从低密度二进制图像到高密度二进制图像的所有二进制图像依次执行该处理，以生成多值图像。 对象多值图像进行密度梯度平滑处理。 结果，获得了放大的多值图像。

公开(公告)号：US06214649B1

公开(公告)日：2001-04-10

申请号：US09373998

申请日：1999-08-16

Applicant: Yoshihiro Ishida

Inventor： Yoshihiro Ishida

IPC: H01L2144

CPC classification number: H01L21/563 ,  H01L24/28 ,  H01L2224/16225 ,  H01L2224/73203 ,  H01L2224/73204 ,  H01L2224/83102 ,  H01L2224/92125 ,  H01L2924/01004 ,  H01L2924/01005 ,  H01L2924/01006 ,  H01L2924/01033 ,  H01L2924/01042 ,  H01L2924/14 ,  H01L2924/15311

Abstract: An IC chip is mounted on a circuit substrate by the flip chip mounting. A projection is provided for forming an eccentric space between the IC chip and the circuit substrate at a position deviated from a center of the IC chip. A sealing resin is injected in a space between the IC chip and the circuit substrate.

Abstract translation: 通过倒装芯片安装将IC芯片安装在电路基板上。 提供突起，用于在偏离IC芯片的中心的位置处在IC芯片和电路基板之间形成偏心空间。 在IC芯片和电路基板之间的空间中注入密封树脂。

公开(公告)号：US5748777A

公开(公告)日：1998-05-05

申请号：US377509

申请日：1995-01-24

Applicant: Akihiro Katayama ,  Yoshihiro Ishida ,  Junichi Yamakawa

Inventor： Akihiro Katayama ,  Yoshihiro Ishida ,  Junichi Yamakawa

IPC: G06K9/48 ,  G06K9/50 ,  G06T5/00 ,  G06T9/20 ,  G06K9/52

CPC classification number: G06K9/481 ,  G06K9/50 ,  G06T7/0083 ,  G06T9/20 ,  G06T2207/10008 ,  G06T2207/30176

Abstract: An outline extraction method and apparatus which extracts a line of interest from image data in order of raster scanning and detects outline vectors of the pixels in the line of interest. The detected outline vectors are sequentially stored in a memory, a state of connection between the stored outline vectors and the other outline vectors in the line of interest is determined, and the outline of the image data is extracted based on the state. Accordingly, coordinates of an intersecting point of the outline of the image data with the coordinate axis of the outline vector is obtained and the image data is coded by generating coded data indicating a white run length and black run length for every line from the coordinates of the intersecting point.

Abstract translation: 一种轮廓提取方法和装置，其以光栅扫描的顺序从图像数据中提取感兴趣的行，并检测所关注线中的像素的轮廓向量。 检测到的轮廓矢量被顺序地存储在存储器中，确定存储的轮廓矢量与感兴趣线中的其它轮廓矢量之间的连接状态，并根据状态提取图像数据的轮廓。 因此，获得图像数据的轮廓与轮廓矢量的坐标轴的交点的坐标，并且通过从坐标系的坐标生成表示每行的白色行程长度和黑色行程长度的编码数据来编码图像数据 相交点。

公开(公告)号：US5721884A

公开(公告)日：1998-02-24

申请号：US312931

申请日：1994-09-30

Applicant: Junichi Shishizuka ,  Yoshinobu Mita ,  Yoshihiro Ishida ,  Miyuki Enokida

Inventor： Junichi Shishizuka ,  Yoshinobu Mita ,  Yoshihiro Ishida ,  Miyuki Enokida

IPC: G06F13/42 ,  G06T11/00 ,  G09G5/39 ,  G06F19/02

CPC classification number: G06T11/001 ,  G06F13/4243 ,  G09G5/39 ,  G09G2360/123 ,  G09G2360/127

Abstract: In a memory-to-memory data transfer apparatus, each of a plurality of memories connected to a memory bus is sequentially selected, and data which occupy predetermined positions in the data read from the selected memory are combined into a set of data. The set of data is sequentially written while addresses in the remaining memory is being updated. Alternatively, a plurality of sets of data are sequentially read out while addresses in a memory are being updated, and data which occupy predetermined positions in the read data are combined into a set of data. The set of data is written while the remaining memories are being sequentially selected. The apparatus may have a plurality of memories connected to a data bus and address generators for generating unique addresses for each of the memories in accordance with a predetermined synchronizing signal. The apparatus, in preferred form, effects transfer of data in thinned-out form or phase-divided form by shifting a generated address or by adding a predetermined value to the address.

Abstract translation: 在存储器到存储器数据传送装置中，顺序地选择连接到存储器总线的多个存储器中的每一个，并且占据从所选存储器读取的数据中的预定位置的数据被组合成一组数据。 在更新剩余存储器中的地址的同时顺序写入数据集。 或者，在更新存储器中的地址的同时顺序地读出多组数据，并且占据读取数据中的预定位置的数据被组合成一组数据。 在依次选择剩余存储器的同时写入数据集。 该装置可以具有连接到数据总线和地址发生器的多个存储器，用于根据预定的同步信号为每个存储器生成唯一的地址。 该装置以优选形式通过移位所生成的地址或通过向地址添加预定值来影响以稀疏形式或相位分割形式的数据传输。

公开(公告)号：US5692210A

公开(公告)日：1997-11-25

申请号：US751560

申请日：1996-11-18

Applicant: Yoshinobu Mita ,  Miyuki Enokida ,  Yoshihiro Ishida ,  Naoto Kawamura

Inventor： Yoshinobu Mita ,  Miyuki Enokida ,  Yoshihiro Ishida ,  Naoto Kawamura

IPC: G06F15/80 ,  G06T1/20 ,  G06F13/00

CPC classification number: G06T1/20 ,  G06F15/8023

Abstract: An image processing apparatus is composed of an image memory comprising a plurality of memory elements, and a processor unit comprising a plurality of processor elements. By suitably engineering the arrangement of the image memory of memory elements, the arrangement of the processor unit of processor elements and the connections among the memory elements and processor elements, it is possible to realize, through a simple construction, high-speed image processing such as image density conversion, image color conversion, image masking, image spatial filtering, image enlargement, image reduction, image rotation, image compression, image extension and image color correction. Also provided are an image processing apparatus for a color image, composed of a plurality of image memories and a plurality of processor elements, and a generalized, multifunctional data parallel processing apparatus capable of processing a large volume of data at high speed.

Abstract translation: 图像处理装置由包括多个存储元件的图像存储器和包括多个处理器元件的处理器单元组成。 通过适当地设计存储元件的图像存储器的布置，处理器元件的处理器单元的布置以及存储元件和处理器元件之间的连接，可以通过简单的结构实现高速图像处理 作为图像浓度转换，图像颜色转换，图像掩蔽，图像空间滤波，图像放大，图像缩小，图像旋转，图像压缩，图像扩展和图像颜色校正。 还提供了一种用于由多个图像存储器和多个处理器元件组成的彩色图像的图像处理装置，以及能够高速处理大量数据的广义的多功能数据并行处理装置。

公开(公告)号：US5515179A

公开(公告)日：1996-05-07

申请号：US171176

申请日：1993-12-22

Applicant: Junichi Yamakawa ,  Yoshihiro Ishida

Inventor： Junichi Yamakawa ,  Yoshihiro Ishida

IPC: G09G5/36 ,  G06T3/00 ,  G06T3/40 ,  G06T9/20 ,  H04N1/40

CPC classification number: G06T9/20

Abstract: Outline vectors representing an arrangement of pixels of a binary image obtained by a binary image obtaining unit are extracted by an outline extractor. From the extracted outline vector, isolated points are extracted. These isolated points are stored in an isolated point data generator 7 as information indicative of the position of isolated point. The outline vectors other than the isolated points are stored in the vector generator 3 as vector data. When a binary image is reproduced, the stored position information of the isolated points is developed to produce vector data by the isolated point data developer 8. Accordingly, since the isolated points are stored not in the form of outline vectors, but position information, the storage capacity for storing the outline vectors can be small. This construction is effective in halftone image in particular.

Abstract translation: 表示由二值图像获取单元获得的二进制图像的像素排列的轮廓矢量由轮廓提取器提取。 从提取的轮廓矢量中提取孤立点。 这些隔离点被存储在隔离点数据生成器7中，作为指示孤立点的位置的信息。 除了孤立点之外的轮廓矢量作为向量数据存储在矢量生成器3中。 当再现二进制图像时，隔离点的存储位置信息被开发以通过孤立点数据显影器8产生矢量数据。因此，由于孤立点不是以轮廓矢量的形式存储，而是存储位置信息， 用于存储轮廓矢量的存储容量可以很小。 这种结构特别适用于半色调图像。

公开(公告)号：US5233225A

公开(公告)日：1993-08-03

申请号：US752172

申请日：1991-08-23

Applicant: Yoshihiro Ishida ,  Katsuji Komatsu ,  Seiichi Mimura ,  Kikuo Takenouchi ,  Isao Yabe ,  Shingo Ichikawa ,  Yoshihiro Shimada

Inventor： Yoshihiro Ishida ,  Katsuji Komatsu ,  Seiichi Mimura ,  Kikuo Takenouchi ,  Isao Yabe ,  Shingo Ichikawa ,  Yoshihiro Shimada

IPC: H01L21/56 ,  H01L23/433

CPC classification number: H01L21/565 ,  H01L23/4334 ,  H01L2224/48227 ,  H01L24/48 ,  H01L2924/00014 ,  H01L2924/01078 ,  H01L2924/1815

Abstract: A resin encapsulated pin grid array includes an IC chip mounted on a resin substrate having a plurality of contact pins on its lower surface and resin-encapsulated by injection molding. A metal heat radiating member is integrally formed on the upper surface of the encapsulating resin when the encapsulating resin is injection-molded.

Abstract translation: 树脂封装的针格栅阵列包括安装在树脂基板上的IC芯片，其下表面具有多个接触针，并通过注射成型树脂封装。 当封装树脂被注射成型时，金属散热构件一体地形成在封装树脂的上表面上。

公开(公告)号：US5162923A

公开(公告)日：1992-11-10

申请号：US312755

申请日：1989-02-21

Applicant: Tadashi Yoshida ,  Naoto Kawamura ,  Yoshihiro Ishida ,  Yasuji Hirabayashi ,  Mitsuru Maeda

Inventor： Tadashi Yoshida ,  Naoto Kawamura ,  Yoshihiro Ishida ,  Yasuji Hirabayashi ,  Mitsuru Maeda

IPC: H04N1/41

CPC classification number: H04N1/41

Abstract: An image encoding method, and an apparatus therefor, enabling rapid comprehension of the outline of the entire image promptly, at the decoding of the encoded image information, and also an image encoding method, and an apparatus therefor, capable of efficient encoding adapted for the state of the image to be encoded. Image information is divided into plural block, each consisting of plural pixels, the frequency of the image in each block is analyzed, the frequency components of the image in each block are separated into plural bands, and the frequency components contained in each of said plural bands are encoded.

Abstract translation: 一种图像编码方法及其装置，能够在对编码图像信息进行解码的同时迅速地理解整个图像的轮廓，以及图像编码方法及其装置，其能够进行有效编码， 要编码的图像的状态。 图像信息被分成多个块，每个块由多个像素组成，每个块中的图像的频率被分析，每个块中的图像的频率分量被分成多个频带，并且包含在所述多个 频带被编码。