Geometric figure position and orientation detection method and device
    1.
    发明授权
    Geometric figure position and orientation detection method and device 失效
    几何图形位置和方位检测方法及装置

    公开(公告)号:US4776027A

    公开(公告)日:1988-10-04

    申请号:US699597

    申请日:1985-02-08

    CPC分类号: G06K9/52 G05D1/0246

    摘要: The position and orientation of a geometric figure are detected by sequentially producing the coordinates with respect to an orthogonal coordinate system of pixels which make up an image of the geometric figure, by singling out from these sequentially produced coordinates of these pixels the coordinates of certain characteristic pixels of these pixels which make up an image of the geometric figure, and by comparing these coordinates of these certain characteristic pixels with stored information. Preferably, the certain characteristic pixels are: that pixel, of those which have the least value of one coordinate, which has the least value of the other coordinate; that pixel, of those which have the greatest value of the one coordinate, which has the least value of the other coordinate; that pixel, of those which have the least value of the other coordinate, which has the least value of the one coordinate; and that pixel, of those which have the greatest value of the other coordinate, which has the least value of the one coordinate. A device is also disclosed for performing this method.

    摘要翻译: 通过从构成几何图像的图像的像素的正交坐标系中顺序产生坐标来检测几何图形的位置和取向,通过从这些像素的这些顺序产生的坐标中选出特定特征的坐标 这些像素的像素构成几何图形的图像,并且将这些特定像素的这些坐标与存储的信息进行比较。 优选地,某些特征像素是:具有最小值的一个坐标的那个像素具有另一个坐标的最小值; 那个具有最大值的一个坐标的那个像素具有另一个坐标的最小值; 那些具有最小值的另一个坐标值的那个像素具有最小的一个坐标值; 以及那个具有最小值的另一个坐标值的那个像素,它具有最小的一个坐标值。 还公开了一种用于执行该方法的装置。

    Illumination system of the digital control type
    2.
    发明授权
    Illumination system of the digital control type 失效
    数字控制型照明系统

    公开(公告)号:US4703344A

    公开(公告)日:1987-10-27

    申请号:US843390

    申请日:1986-03-24

    摘要: A digital control color illumination system is disclosed for use in a visual system of an industrial robot or the like. This system comprises: image memories to store the color data regarding illumination patterns of three primary colors of red, green, and blue; video signal producing circuits each for reading out the data from the image memory and converting it to the video signal indicative of the illumination pattern; video projectors each for projecting the light of the illumination pattern represented by the video signal onto an object to be illuminated; and a color TV camera to pick up the object image. The image data picked up is processed by a computer system to produce the data which is stored into the image memories. With this illumination system, the scanning of the illumination patterns, and the change, adjustment, storage, and the like of the illumination patterns can be executed by software in accordance with programs preliminarily stored in the program memory in the computer system.

    摘要翻译: 公开了用于工业机器人等的视觉系统中的数字控制彩色照明系统。 该系统包括:存储有关红色,绿色和蓝色三原色的照明图案的颜色数据的图像存储器; 每个视频信号产生电路用于从图像存储器读出数据并将其转换成指示照明图案的视频信号; 每个用于将由视频信号表示的照明图案的光投射到被照明物体上的视频投影仪; 和彩色电视摄像机来拾取物体图像。 拾取的图像数据由计算机系统处理以产生存储到图像存储器中的数据。 利用该照明系统,可以通过软件根据预先存储在计算机系统中的程序存储器中的程序来执行照明图案的扫描以及照明图案的改变,调整,存储等。

    System for recognizing three-dimensional objects
    3.
    发明授权
    System for recognizing three-dimensional objects 失效
    用于识别三维物体的系统

    公开(公告)号:US4654872A

    公开(公告)日:1987-03-31

    申请号:US634043

    申请日:1984-07-24

    IPC分类号: G06T7/00 H04N13/00 G06K9/00

    摘要: A system for recognizing a three-dimensional object includes a plurality of image pickup apparatus, e.g. TV cameras, by which images of the object are picked up from at least three directions. Feature points are extracted from each of at least three images obtained. Two sets of feature points on epipolarly lines are formed which lines are formed on at least two of the images by a feature point on another image. A set of feature points is selected which satisfy a restricting condition determined by the relationship of the image picking-up direction.

    摘要翻译: 用于识别三维物体的系统包括多个图像拾取装置,例如, 从至少三个方向拾取物体的图像的电视摄像机。 从获得的至少三个图像中的每一个提取特征点。 形成在双极线上的两组特征点,其中通过另一图像上的特征点在至少两个图像上形成这些线。 选择满足由图像拾取方向的关系确定的限制条件的一组特征点。

    Pattern outline tracking method and apparatus
    4.
    发明授权
    Pattern outline tracking method and apparatus 失效
    模式轮廓跟踪方法和装置

    公开(公告)号:US4703512A

    公开(公告)日:1987-10-27

    申请号:US760607

    申请日:1985-07-30

    CPC分类号: G06T9/20 G06K9/48

    摘要: A pattern outline tracking system provides the functions that include scanning an object to be recognized and providing a series of analog video signals, comparing a video signal level with the specific, variable threshold level and providing a binary signal consisting of a sequence of serial bits, holding the binary signal bits as the input image pattern in an image memory, deriving partial patterns sequentially from the serial binary signal bits, each of the partial patterns consisting of a matrix of vertical and horizontal lines each containing specific numbers of pixels, matching the partial patterns against the corresponding reference patterns containing the previously defined starting points to be tracked, holding the address in the image memory for the certain pixel within the partial pattern identified in the matching step as the possible point address for the tracking operation, and tracking the outline of the input image pattern on the image memory in accordance with the possible point address.

    摘要翻译: 模式轮廓跟踪系统提供了包括扫描要识别的对象并提供一系列模拟视频信号的功能,将视频信号电平与特定的可变阈值电平进行比较,并提供由串行位序列组成的二进制信号, 将二进制信号位保持为图像存储器中的输入图像图案,从串行二进制信号比特顺序地导出部分图案,每个部分图案由垂直和水平线的矩阵组成,每个包含特定数量的像素,匹配部分 针对包含先前定义的要跟踪的起始点的相应参考图案的图案,将在匹配步骤中标识的部分图案内的特定像素的图像存储器中的地址保持为跟踪操作的可能点地址,并跟踪轮廓 的图像存储器中的输入图像图案 可点地址。

    Pattern analyzer
    5.
    发明授权
    Pattern analyzer 失效
    图案分析仪

    公开(公告)号:US4601057A

    公开(公告)日:1986-07-15

    申请号:US559922

    申请日:1983-12-09

    CPC分类号: G06K9/38

    摘要: A pattern analyzer includes a sensor (19) for picking up an image of a pattern at a plurality of sampling points and for producing a shade signal representing the depth of shade at each sampling point. The sampled shade signals are temporarily stored in a memory (23) and, at the same time, are classified into different shade depths by a histogram forming circuit (30, 31, 32 and 33). The result of the classification is such that an amount of classified shade signals reaches a peak at two different shade depths. A threshold level determining circuit (34, 35, 36 and 37) is provided for determining a threshold level of shade depth between the two different shade depths. Also, a comparator (38, 39, 40 and 41) is provided for comparing the shade signals stored in the memory with the threshold level and for producing a 1-bit binary signal for each shade signal.

    摘要翻译: 图案分析器包括用于拾取多个采样点处的图案的图像的传感器(19),并且用于产生表示每个采样点的阴影深度的色调信号。 采样的阴影信号被临时存储在存储器(23)中,并且同时被直方图形成电路(30,31,32和33)分类为不同的色深。 分类结果使得分类阴影信号的数量在两个不同色调深度处达到峰值。 提供阈值水平确定电路(34,35,36和37),用于确定两个不同色调深度之间的色调深度的阈值水平。 此外,提供比较器(38,39,40和41),用于将存储在存储器中的色调信号与阈值电平进行比较,并用于为每个色调信号产生1位二进制信号。