首页> 外国专利> IMAGE FEATURE POINT EXTRACTION METHOD, DEFECT INSPECTION METHOD, AND DEFECT INSPECTION DEVICE

IMAGE FEATURE POINT EXTRACTION METHOD, DEFECT INSPECTION METHOD, AND DEFECT INSPECTION DEVICE

机译:图像特征点提取方法,缺陷检查方法以及缺陷检查装置

摘要

PROBLEM TO BE SOLVED: To provide an image feature point extraction method for capturing a grayscale image to be inspected as an input and obtaining the coordinate value of a (conspicuous) area having a visually clear feature in the image and an index value for clearness (degree of conspicuousness), and a defect inspection method and defect inspection device for determining the presence of a defect on the basis of the index value.;SOLUTION: A total gradation value of one selected pixel and pixels in its periphery, n0 pixels in total, is calculated with respect to all pixels of a given grayscale image. Also, probability distribution of gradation values of all pixels present inside a sufficiently wide area including those pixels is calculated, and the sum of probability distribution in a total of gradation values of the n0 pixels is made an index of appearance probability of a pixel of interest. When this value is smaller than a designated threshold, a (conspicuous) pixel having a visually clear feature and its peripheral region are assumed.;COPYRIGHT: (C)2015,JPO&INPIT
机译:解决的问题:提供一种图像特征点提取方法,用于捕获要检查的灰度图像作为输入,并获取图像中具有视觉上清晰特征的(显着)区域的坐标值和清晰度指标(显着程度),以及一种基于指标值确定缺陷是否存在的缺陷检查方法和缺陷检查设备;解决方案:一个选定像素及其周围像素的总灰度值,总共n0个像素是针对给定灰度图像的所有像素计算的。此外,计算在包括那些像素的足够宽的区域内存在的所有像素的灰度值的概率分布,并且将n0个像素的灰度值的总和中的概率分布之和作为关注像素的出现概率的指标。 。当此值小于指定的阈值时,将假定具有视觉清晰特征的(显眼)像素及其外围区域。版权所有:(C)2015,JPO&INPIT

著录项

  • 公开/公告号JP2015135540A

    专利类型

  • 公开/公告日2015-07-27

    原文格式PDF

  • 申请/专利权人 DAINIPPON PRINTING CO LTD;

    申请/专利号JP20140005670

  • 发明设计人 KOBAYASHI HIDEAKI;

    申请日2014-01-16

  • 分类号G06T7/00;G06T1/00;G01N21/88;

  • 国家 JP

  • 入库时间 2022-08-21 15:34:24

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号