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
    • 4. 发明申请
    • IMAGE ENCODER, IMAGE DECODER, IMAGE ENCODING METHOD, IMAGE DECODING METHOD AND IMAGE ENCODING/DECODING SYSTEM
    • 图像编码器,图像解码器,图像编码方法,图像解码方法和图像编码/解码系统
    • WO1998042135A1
    • 1998-09-24
    • PCT/JP1997003825
    • 1997-10-23
    • MITSUBISHI DENKI KABUSHIKI KAISHASEKIGUCHI, ShunichiASAI, KohtaroMURAKAMI, TokumichiNISHIKAWA, HirofumiKURODA, ShinichiISU, YoshimiHASEGAWA, Yuri
    • MITSUBISHI DENKI KABUSHIKI KAISHA
    • H04N07/32
    • H04N19/523H04N19/53H04N19/537H04N19/61
    • An image encoder and an image decoder which perform the image displacement other than parallel displacement on a relatively small scale and with which the processing time can be reduced and estimation with a high precision is enabled. The constitution for compression-coding the input image comprises a movement compensating means which detects the frame-to-frame movement of the blocks into which the input image is divided. The constitution further comprises a movement detection unit which includes a deformed block matching section which deforms only integer pixels, i.e. real sampling points present in a partial region corresponding to a reference image for movement detection into predetermined formats, cuts the deformed pixels out, and compares the cut-out deformed pixels with the integer pixels of the blocks of the input image and which outputs a moving vector which gives a minimum error extracted by the coordinate designation and a movement compensation unit which includes a corresponding point determination section which includes the deformed block matching section, correlates the blocks of the reference image in accordance with movement parameters obtained from comparison output, designates the coordinates, deforms and determines the block, and which outputs a predicted partial image. Further, an image decoder is so constructed as to correspond to the construction of the encoder. Moreover, the image decoder calculates the coordinates with a plurality of movement vectors by using also the coordinate values of the half-pixels of the reference image in a movement compensating means and subjects the obtained pixel values to a deformation processing by using a deformation pattern in accordance with a deformed pattern.
    • 执行图像编码器和图像解码器,该图像编码器和图像解码器能够以相对较小的尺度进行除了平行位移之外的图像位移,并且能够降低处理时间并且能够以高精度进行估计。 用于对输入图像进行压缩编码的结构包括移动补偿装置,其检测输入图像被划分成的块的帧到帧的移动。 该结构还包括运动检测单元,其包括变形块匹配部分,该变形块匹配部分仅使整数像素变形,即,存在于与用于运动检测的参考图像相对应的预定格式的部分区域中的实际采样点,将变形像素切出,并且比较 切出的变形像素与输入图像的块的整数像素,并输出给出由坐标指定提取的最小误差的运动矢量,以及运动补偿单元,其包括对应点确定部分,该对应点确定部分包括变形块 匹配部分,根据从比较输出获得的移动参数相关联参考图像的块,指定坐标,变形并确定块,并且输出预测的部分图像。 此外,图像解码器被构造成对应于编码器的结构。 此外,图像解码器还通过使用移动补偿装置中的参考图像的半像素的坐标值来计算具有多个运动矢量的坐标,并且通过使用变形模式对所获得的像素值进行变形处理 根据变形的图案。
    • 5. 发明申请
    • IMAGE ENCODER, IMAGE DECODER, IMAGE ENCODING METHOD, IMAGE DECODING METHOD AND IMAGE ENCODING/DECODING SYSTEM
    • 图像编码器,图像解码器,图像编码方法,图像解码方法和图像编码/解码系统
    • WO1998042134A1
    • 1998-09-24
    • PCT/JP1997000834
    • 1997-03-17
    • MITSUBISHI DENKI KABUSHIKI KAISHASEKIGUCHI, ShunichiASAI, KohtaroMURAKAMI, Tokumichi
    • MITSUBISHI DENKI KABUSHIKI KAISHA
    • H04N07/32
    • H04N19/523H04N19/53H04N19/537H04N19/61
    • An image encoder and an image decoder which facilitate the prediction of the image movement other than the parallel movement with a relatively small scale process, with a shortened processing time and with a high precision. An input image is divided into predetermined blocks and the input image is compressed and encoded by providing a means which predicts the movement compensation by detecting the movements between the frames of the blocks. A movement detection unit which includes a deformed block matching section which deforms only integer pixels which are actual sample points existing in partial regions corresponding to a reference image for the movement detection into required shapes and segments them and compares the deformed and segmented pixels with the integer pixels of the blocks of the input image and which outputs a motion vector which gives a minimum error which is extracted by designating the coordinates and a movement compensation unit which includes a corresponding point determining section which specifies the coordinates of the integer pixels correspondingly to the blocks of the reference image in accordance with the movement parameters which are obtained from the comparison output including the output from the deformed block matching section and deforms the pixels and determines and which outputs a predicted partial image are provided. The image decoder is so constructed as to correspond to the image encoder.
    • 图像编码器和图像解码器,其具有缩短的处理时间和高精度,有助于以比较小规模的处理来预测平行移动以外的图像运动。 输入图像被划分为预定块,并且通过提供通过检测块的帧之间的移动来预测移动补偿的装置来对输入图像进行压缩和编码。 一种移动检测单元,其包括变形块匹配部分,该变形块匹配部分仅将存在于与用于运动检测的参考图像相对应的部分区域中的实际采样点变形为所需形状的整数像素,并将其分段,并将变形和分割的像素与整数 输出图像的块的像素,并输出给出通过指定坐标提取的最小误差的运动矢量,以及移动补偿单元,该运动补偿单元包括对应点确定部,该对应点确定部指定与块相对应的整数像素的坐标 根据从包括来自变形块匹配部分的输出的比较输出获得的运动参数,并且使像素变形并且确定并输出预测的部分图像。 图像解码器被构造成对应于图像编码器。
    • 6. 发明申请
    • PULSE POWER SOURCE APPARATUS
    • 脉冲电源设备
    • WO1999034502A1
    • 1999-07-08
    • PCT/JP1997004849
    • 1997-12-25
    • MITSUBISHI DENKI KABUSHIKI KAISHAYABUUCHI, MasatakaIWATA, Akihiko
    • MITSUBISHI DENKI KABUSHIKI KAISHA
    • H02M09/04
    • H03K17/80H03K3/53H05B41/34
    • A pulse power source apparatus which quickly discharges the charged voltage of a capacitor (23) by closing a switch (24). First and second serial circuits (51 and 52) are connected in series between the output terminals of a DC voltage source (2). The first serial circuit (51) is composed of a first reactor (14) and a first forward diode (15), and the second serial circuit (52) is composed of a second forward diode (16) and the switch (24). A third serial circuit (53) composed of the capacitor (23), a second reactor (25), and a discharge tube (9), and the discharge tube is connected in parallel with the second serial circuit (52). When the switch (24) is closed, the discharge tube (9) emits light, but the discharged current from the capacitor (23) becomes an oscillating current and a voltage of the opposite polarity to that of the voltage generated at the initial charging time is generated in the capacitor (23). Since the oscillating voltage is superposed upon the power supply voltage and utilized at the next charging cycle time, an extremely efficient pulse power source apparatus can be obtained.
    • 一种通过闭合开关(24)来快速放电电容器(23)的充电电压的脉冲电源装置。 第一和第二串行电路(51和52)串联连接在DC电压源(2)的输出端子之间。 第一串联电路(51)由第一电抗器(14)和第一正二极管(15)组成,第二串联电路(52)由第二正向二极管(16)和开关(24)构成。 由电容器(23),第二反应器(25)和放电管(9)组成的第三串联电路(53),放电管与第二串联电路(52)并联连接。 当开关(24)关闭时,放电管(9)发光,但是来自电容器(23)的放电电流变为振荡电流,并且与初始充电时间产生的电压相反极性的电压 在电容器(23)中产生。 由于振荡电压叠加在电源电压上并在下一个充电周期时间被利用,所以可以获得极其有效的脉冲电源装置。