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
    • 1. 发明授权
    • Cascaded arithmetic pipeline data processor
    • 级联算术流水线数据处理器
    • US6154829A
    • 2000-11-28
    • US174562
    • 1998-10-19
    • Yoshiteru MinoTadashi OkamotoHiroshi Kadota
    • Yoshiteru MinoTadashi OkamotoHiroshi Kadota
    • G06F9/38G06F7/38
    • G06F9/3861G06F9/3867G06F9/3869G06F9/3875
    • Five processing units, namely one data memory, three arithmetic units, and one data memory, are connected together in a cascade arrangement so as to form a single arithmetic pipeline. Likewise, five control devices are connected together in a cascade arrangement and a control signal requesting that a series of data processing operations should start is sent to the first stage control device. Each control device starts to send a micro instruction to a corresponding processing unit upon detection of a processing start request bit in the received control signal and sends a signal which lags the control signal by a delay time equal to a number of cycles required to complete a processing operation of the processing unit, to the next stage control device. The first stage control device is provided with a loop counter operable to count the number of times processing is repeated and automatically generates a processing start request and a processing end request to the next stage control device.
    • 五个处理单元,即一个数据存储器,三个运算单元和一个数据存储器,以级联布置连接在一起,以形成单个运算管线。 类似地,五个控制装置以级联布置连接在一起,并且要求将一系列数据处理操作开始的控制信号发送到第一级控制装置。 每个控制装置在检测到所接收的控制信号中的处理开始请求位后,开始向对应的处理单元发送微指令,并发送一个延迟时间等于完成一个控制信号所需周期的延迟时间的信号 处理单元的处理操作,到下一级控制装置。 第一级控制装置设置有循环计数器,用于对重复处理的次数进行计数,并且自动地向下一级控制装置生成处理开始请求和处理结束请求。
    • 2. 发明授权
    • Reconfigurable digit-serial arithmetic system having a plurality of digit-serial arithmetic units
    • 具有多个数字串行运算单元的可重构数字串行运算系统
    • US06230175B1
    • 2001-05-08
    • US09188388
    • 1998-11-10
    • Tadashi OkamotoHiroshi KadotaYoshiteru Mino
    • Tadashi OkamotoHiroshi KadotaYoshiteru Mino
    • G06F1531
    • G06F7/5324G06F7/49994G06F7/504G06F7/57G06F2207/386
    • A bus for data transmission, bus switches for slicing the bus, and four arithmetic blocks are provided to perform a series of fixed-point arithmetic operations. Each of the four arithmetic blocks has a plurality of digit-serial arithmetic units, namely a multiplier, an adder/subtracter, and a shifter. Each of the digit-serial arithmetic units has the functions of receiving a plurality of input digits representative of an input operand and a digit position indicator of each of the input digits and providing a plurality of result digits representative of an arithmetic result of the input operand and a digit position indicator of each of the result digits to any other one of the plurality of digit-serial arithmetic units. Particularly, the digit-serial adder/subtracter contains therein a selector for input switching so that the digit-serial adder/subtracter can perform butterfly arithmetic.
    • 提供用于数据传输的总线,用于分割总线的总线开关和四个运算块,以执行一系列定点算术运算。 四个算术块中的每一个具有多个数字串行运算单元,即乘法器,加法器/减法器和移位器。 每个数字串行运算单元具有接收表示每个输入数字的输入操作数和数字位置指示符的多个输入数字的功能,并提供表示输入操作数的算术结果的多个结果数字 以及每个结果数字的数字位置指示符到多个数字串行运算单元中的任何另一个。 特别地,数字串联加法器/减法器包含用于输入切换的选择器,使得数字串行加法器/减法器可以执行蝴蝶运算。
    • 3. 发明授权
    • Data processor
    • 数据处理器
    • US06125438A
    • 2000-09-26
    • US63009
    • 1998-04-21
    • Tadashi OkamotoHiroshi KadotaYoshiteru Mino
    • Tadashi OkamotoHiroshi KadotaYoshiteru Mino
    • G06F15/78G06F15/00
    • G06F15/7857G06F15/7864
    • A data processor of the invention includes plural memories, plural arithmetic units, a data transfer unit and a network. The data transfer unit transfers various data to predetermined memories, and switches the connections between the memories and the arithmetic units by using the network. The control unit adds a processability judgement signal to a data read from a predetermined memory in reading the data, so as to make a pair of the data and the processability judgement signal. Each of the arithmetic units receives the data and the processability judgement signal, conducts predetermined processing on the received data, delays the received processability judgement signal by the number of cycles equal to its own processing cycle, and outputs resultant data obtained through the processing and the delayed processability judgement signal. Accordingly, in storing ultimate resultant data in a predetermined storage unit, the storage unit stores the ultimate resultant data as effective data on the basis of the processability judgement signal added to the resultant data. The data processor attains wide application and can be applied to various types of multimedia applications by switching the connections between the memories and the arithmetic units by using the network.
    • 本发明的数据处理器包括多个存储器,多个运算单元,数据传送单元和网络。 数据传送单元将各种数据传送到预定存储器,并且通过使用网络来切换存储器和算术单元之间的连接。 控制单元在读取数据时,对从预定存储器读取的数据添加可处理性判断信号,以便形成一对数据和加工性判定信号。 每个算术单元接收数据和可处理性判断信号,对接收到的数据进行预定的处理,将接收到的加工性判断信号延迟等于其自身的处理周期的周期数,并输出通过处理获得的结果数据和 延迟加工性判断信号。 因此,在将最终结果数据存储在预定存储单元中时,存储单元基于加到结果数据的加工性判断信号将最终结果数据存储为有效数据。 数据处理器具有广泛的应用,可以通过使用网络切换存储器和算术单元之间的连接来应用于各种类型的多媒体应用。
    • 4. 发明授权
    • Image processing apparatus and image processing method for favorably
enhancing continuous boundaries which are affected by noise
    • 图像处理装置和图像处理方法,有利于增强受噪声影响的连续边界
    • US5754618A
    • 1998-05-19
    • US772756
    • 1996-12-23
    • Tadashi OkamotoYoshiteru MinoHiroshi Kadota
    • Tadashi OkamotoYoshiteru MinoHiroshi Kadota
    • G06T5/20G06K9/00
    • G06T5/20A61B8/13A61B8/483Y10S378/901
    • An image processing apparatus which performs a process which enhances incongruous pixel values in an original image, including an object pixel selecting unit for selecting an object pixel out of pixels which compose the original image, a plurality n of filter units, each of which extracts pixel values of a set of pixels, including the selected object pixel, on one of a predetermined surface and a predetermined line which pass through the object pixel, out of pixels in a predetermined range, wherein each predetermined surface and predetermined line is at a different inclination, a relative size comparing unit for comparing pixels values of each set of pixels extracted by each filter unit and specifying a filter unit whose pixel values best approximate to a predetermined standard, and a first pixel value enhancing unit for enhancing a pixel value of the object pixel based on only the pixel values of the specified filter unit.
    • 一种图像处理装置,其执行增强原始图像中的不协调像素值的处理,包括用于选择构成原始图像的像素中的对象像素的对象像素选择单元,其中每个滤色器单元提取像素 在预定范围内的像素中,在预定表面和穿过对象像素的预定线之一上的包括所选择的对象像素的一组像素的值,其中每个预定表面和预定线处于不同的倾斜度 相对尺寸比较单元,用于比较由每个滤波器单元提取的每组像素的像素值,并指定最佳近似于预定标准的像素值的滤波器单元;以及第一像素值增强单元,用于增强对象的像素值 仅基于指定滤波器单元的像素值的像素。
    • 6. 发明授权
    • Address conversion apparatus
    • 地址转换装置
    • US4910668A
    • 1990-03-20
    • US100561
    • 1987-09-24
    • Tadashi OkamotoHiroshi KadotaMasaitsu Nakajima
    • Tadashi OkamotoHiroshi KadotaMasaitsu Nakajima
    • G06F12/10G06F12/12G11C15/04
    • G06F12/1036G06F12/123
    • An address conversion apparatus includes a content addressable memory for storing a plurality of logical addresses, and a random access memory for storing a plurality of physical addresses corresponding to the logical addresses. When an input logical address is received, a search is conducted to find the same logical address stored in the memory. When the same logical address is found, the content addressable memory causes the random access memory to output a corresponding physical address. The content addressable memory includes a plurality of logical address storage units. Each unit has a plurality of data bit cells for storing address data and a process identification number cell for storing a process identification number. Thereby, a plurality of logical addresses which correspond to different processes are stored in the single content addressable memory.
    • 地址转换装置包括用于存储多个逻辑地址的内容可寻址存储器和用于存储对应于逻辑地址的多个物理地址的随机存取存储器。 当接收到输入逻辑地址时,进行搜索以找到存储在存储器中的相同逻辑地址。 当找到相同的逻辑地址时,内容可寻址存储器使随机存取存储器输出相应的物理地址。 内容可寻址存储器包括多个逻辑地址存储单元。 每个单元具有用于存储地址数据的多个数据位单元和用于存储处理标识号的处理标识号单元。 因此,与单个内容可寻址存储器中存储对应于不同进程的多个逻辑地址。