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
    • 42. 发明专利
    • INSPECTION EQUIPMENT FOR PRINTED MARK
    • JPH06323825A
    • 1994-11-25
    • JP11283993
    • 1993-05-14
    • MITSUBISHI ELECTRIC CORP
    • SAKAGAMI YOSHIKAZUSATO HAJIMEKIMURA YORIAKIYOSHIDA KOSAKU
    • G01B11/24G01B11/26
    • PURPOSE:To decide the abnormality (defect) of a mark accurately and easily by comparing a sample image subjected to correction of inclination angle with an operated sample image. CONSTITUTION:An image picked up by a first device in a device group printed with an identical character and an identical mark is subjected to A/D conversion 2 and stored in a multilevel memory 3. On the other hand, a CPU 10 determines the inclination angle of the image and operates the position of character region. The CPU 10 also operates an optimal threshold value for binarization based on the results of lightness histogram and binarizes the data stored in the memory 3 and then stores the binarized data in a memory 8 for sample image. The image of device in second and subsequent groups is then processed similarly and the inclination angle of a sample image thus operated is corrected 13 by the difference from the inclination angle of a sample image and stored in a sample image memory 6. A difference detector 14 compares the contents of the memories 6, 8 and if any difference is detected, the longitudinal, lateral, and oblique dimensions thereat are detected 15-18. If the detected distance exceeds a predetermined distance, a decision is made that the mark is abnormal and an alarm 19 is delivered.
    • 43. 发明专利
    • IMAGE PROCESSING SYSTEM
    • JPH04205477A
    • 1992-07-27
    • JP33667590
    • 1990-11-30
    • MITSUBISHI ELECTRIC CORP
    • SAKAGAMI YOSHIKAZUKAWATO SHINJIROHASHIMOTO MANABU
    • G06T7/60
    • PURPOSE:To efficiently obtain an objective line from a screen by providing this image processing system with a preprocessor and a start point detecting circuit for applying a range determining condition to a feature value and selecting picture elements included in the range of the condition out of picture elements included within a determine area of an image or the edge points of the whole screen as the candidates of a start point. CONSTITUTION:The image processing system is provided with the preprocessor 3 for determining a picture element whose brightness is changed as an edge point and finding out the intensity of the edge point, a brightness changing direction on the edge point, the brightness of a point near the edge point, and a feature value data of a distance up to the succeeding edge point. The start point detecting circuit 6 forms the determined range of respective feature values obtained by the preprocessing and selects picture elements included in the condition range out of picture elements in a determined area of the screen or edge point picture elements on the whole screen as the candidates of the start point. Since only effective start points can be detected, an operation time required for the selection of the edge points and the division of a connection line can be shortened and the measurement of the size, position, status, etc., of an object or its image recognition can be rapidly and efficiently executed.
    • 44. 发明专利
    • SHADING CORRECTION CIRCUIT
    • JPH01291576A
    • 1989-11-24
    • JP12126088
    • 1988-05-18
    • MITSUBISHI ELECTRIC CORP
    • WASHIMI KAZUHIKOSAKAGAMI YOSHIKAZUHASHIMOTO MANABU
    • H04N5/16
    • PURPOSE:To miniaturize a circuit with a low cost and to perform correction by shading by generating a reference voltage varied by an offset, a gain, and the regulated amount of the shading by extracting a field component, and digitizing an input video signal based on the reference voltage. CONSTITUTION:The input video signal 20 is inputted to a field discrimination circuit 1, and a first field is discriminated from a second field, then, it goes to a field discrimination signal. The signal drives analog switches 2a and 2b, and either gain out of two kinds of gains of a video signal amplifier 5 set in advance by gain control variable resistors 4a and 4b is selected, and either two kinds of offsets set in advance by offset control variable resistors 3a and 3b is selected. The shading between the fields of the gain and the offset are corrected, and a video signal outputted from the video signal amplifier 5 goes to a signal without having the shading between the fields. The video signal is A/D-converted by an A/D converter 6, and a digitized video signal 21 can be obtained.
    • 46. 发明专利
    • DEFECT DETECTING DEVICE
    • JPS6365583A
    • 1988-03-24
    • JP21038386
    • 1986-09-05
    • MITSUBISHI ELECTRIC CORP
    • NAKAJIMA KEIJISAKAGAMI YOSHIKAZU
    • G01N21/88G06T7/00
    • PURPOSE:To reduce a quantizing error and to detect a defect with high accuracy by providing a shift quantity detecting circuit, correcting a shift quantity whenever two pieces of images correspond to each other, and also, executing a calculation up to a decimal part. CONSTITUTION:An image input is obtained as the value of '0' or '1' by allowing it to pass through a binarization circuit 16. If a variation occurs in a position of (X0, Y0) of one image, whether the same variation is occurred in the position of coordinates (X0+dx, Y0+dy) to which shift quantities (dx), (dy) are added, in the other image or not is detected. Unless the same variation is occurred within (k) picture elements centering around its position [(k) picture elements have the smallest size that are decided to be a defect], its position is written as a defect in a memory device B6. If the same variation is detected within the range of + or -1 picture element, the correction of the shift quantity is executed. The shift quantity is derived by collecting 16 pieces of photographed coordinate values of coordinates in which two pieces of images correspond to each other, and averaging them, and at the time of correcting the shift quantity, 16 pieces of the latest data are used, and by deriving this value up to a decimal part, the accuracy is raised.
    • 48. 发明专利
    • Picture recognizing device
    • 图像识别装置
    • JPS6116375A
    • 1986-01-24
    • JP13691384
    • 1984-07-02
    • Mitsubishi Electric Corp
    • SAKAGAMI YOSHIKAZUTSURUSE HIDENORI
    • G06T7/00G06K9/00G06T7/60
    • PURPOSE: To cope with many objects or a complicated shape of the object by dividing brightness change points in each horizontal scanning line to prescribed areas and storing brightness change points in each divided area in a storage means group.
      CONSTITUTION: The video signal from an input terminal 7 is given to D flip- flop 31 through a buffer 9 and a binarizing circuit 10. Since a binary counter 40 and a decoder 41 change individual output lines from the high level or the low level to the low level or the high level each time the binarized video signal is changed from light or darkness to darkness to light, the length to the brightness change point is taken into a latch circuit 45 each time. Length-to-brightness change points are divided to prescribed areas and are taken into latch circuits 45-1W45-3. A computer 46 completes the shape of the object on the basis of stored contents of latch circuits 45.
      COPYRIGHT: (C)1986,JPO&Japio
    • 目的:通过将每个水平扫描线中的亮度变化点划分为规定区域并存储在存储装置组中的每个划分区域中的亮度变化点,来处理物体的许多物体或复杂形状。 构成:来自输入端子7的视频信号通过缓冲器9和二值化电路10被提供给D触发器31.由于二进制计数器40和解码器41将各个输出线从高电平或低电平改变为 每当二值化视频信号从亮度或暗度变化到暗度至亮度时,低电平或高电平,则每次将亮度变化点的长度插入锁存电路45。 长度到亮度的变化点被划分为规定的区域,并被插入到锁存电路45-1-45-3中。 计算机46基于锁存电路45的存储内容完成对象的形状。
    • 50. 发明专利
    • Position measuring device
    • 位置测量装置
    • JPS58217088A
    • 1983-12-16
    • JP10049882
    • 1982-06-11
    • Mitsubishi Electric Corp
    • SAKAGAMI YOSHIKAZUNEKOTSUKA MASAMICHITSUTSUMI HIRONOBUHIRONAKA KAZUMITSUAZUMA KENICHI
    • H04N7/18G06K9/48G06T1/00H01L21/66
    • G06K9/48
    • PURPOSE:To decide accurately the position of a subject, by performing an operation between a standard pattern and the picture signal supplied from a TV camera. CONSTITUTION:The signal obtained from scanning through a TV camera is binary coded and delivered. In this case, the time during which a horizontal synchronizing signal is delivered from the TV camera and then the logic of a video signal is inverted is measured. This time is equivalent to the distance from the edge of a picture to the contour of a subject. A microcomputer checks the correspondence between the time data needed for inversion and the order of the corresponding horizontal synchronizing signal and stores it in a memory. Then the microcomputer measures the length of the straight line segment of a contour line and compares it with the data of a standard subject which is previously supplied. Thus the present position of the subject is decided. For a wire bonder, the accuracy is improved for the subsequent bonding by teaching and showing first a chip to the TV camera.
    • 目的:通过执行标准图案与从电视摄像机提供的图像信号之间的操作来准确地确定被摄体的位置。 构成:通过电视摄像机扫描获得的信号是二进制编码和传送的。 在这种情况下,测量从TV摄像机传送水平同步信号,然后反转视频信号的逻辑的时间。 这个时间相当于从图片边缘到被摄体轮廓的距离。 微机检查反转所需的时间数据与相应的水平同步信号的顺序之间的对应关系,并将其存储在存储器中。 然后微型计算机测量轮廓线的直线段的长度,并将其与先前提供的标准物体的数据进行比较。 因此,决定对象的当前位置。 对于引线键合机,通过教导提高随后的接合精度,并将第一个芯片显示给电视摄像机。