首页> 外国专利> bildbasierte objekterkennungstechnik

bildbasierte objekterkennungstechnik

机译:基于图像的目标识别技术

摘要

a comprehensive software product storage device to perform a method for classification of bildstrukturmerkmalen (b [i]) within a bilddatenelements (2), and the software for installation on a unit of account with access to an imaging sensor is trained to a bilddatenelement (2) to relateand with the bilddatenelement (2) bildinformationswerte (a [i]) with a known grid reference (rr) are stored, and with the bildinformationswerte (a [i]) in a known value range (a wb) occur, characterized by the fact that the executable procedure the following steps comprises: - a first and a second micro area (mb1,mb2) with known grid positions (i) within the bilddatenelements (2) will be selected; the bildinformationswerte (a [i]) in the first and the second micro area (mb1, wertemu00e4u00dfig mb2) are compared (h1, h2, h3, h4), with at least a differential value (d [j]) between the b ildinformationswerten (a [i]) of the first and the second micro bildbereichs (mb1,mb2) is calculated within a possible range, the differenzwerte (differential value range (wb) are at least three on - bereichsabschnitte (d - a1 (a2, d - a3 / db, 1 db 2, d - b3) predetermined; it difference features (dm1, dm2 / dm1, dm2) as results of v on predefined logical statements (l1, l2 and l1 ',l2 ') on the processed bildinformationswerte (a [i])); - the logical statements (l1, l2) allow at least a provision on the on - bereichsabschnitt (d - a1, a2, a3 d, d) a reference differential value (d [j]).
机译:一个综合的软件产品存储设备,用于执行对bilddatenelements(2)中的bildstrukturmerkmalen(b [i])进行分类的方法,并且将安装在可访问成像传感器的帐户单位上的软件训练为bilddatenelement(2)。 )与bilddatenelement(2)相关并且与已知网格参考(rr)一起存储的bildinformationswerte(a [i]),并且在已知值范围(a wb)中与bildinformationswerte(a [i])发生,其特征在于可执行过程包括以下步骤:-将选择在二元元素(2)中具有已知网格位置(i)的第一和第二微区域(mb1,mb2);比较第一个和第二个微区域(mb1,wertem u00e4 u00dfig mb2)中的bildinformationswerte(a [i]),并至少将其微分值(d [j])进行比较(h1,h2,h3,h4)在一个可能的范围内计算出第一和第二个微型bildbereichs的b ildinformationswerten(a [i])(mb1,mb2)之间,diffenzwerte(微分值范围(wb))至少为三个on-bereichsabschnitte(d-a1 (a2,d-a3 / db,1 db 2,d-b3)预先确定;它在预定义逻辑语句(l1,l2和l1',l2)上将v的结果与特征(dm1,dm2 / dm1,dm2)相差')在已处理的bildinformationswerte(a [i])上;-逻辑语句(l1,l2)至少允许在on-bereichsabschnitt(d-a1,a2,a3 d,d)上提供一个参考微分值( d [j])。

著录项

  • 公开/公告号DE202015100437U1

    专利类型

  • 公开/公告日2016-03-31

    原文格式PDF

  • 申请/专利权人 ADASENS AUTOMOTIVE GMBH;

    申请/专利号DE202015100437U1

  • 发明设计人

    申请日2015-01-30

  • 分类号G06K9/62;

  • 国家 DE

  • 入库时间 2022-08-21 14:09:01

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