首页> 外国专利> PATTERN RECOGNIZING DEVICE WITH PATTERN MATCHING IN SLANT PARALLELOGRAMMIC BLOCKS OF WIDTHS DEPENDENT ON CLASSIFIED REFERENCE PATTERN LENGTHS

PATTERN RECOGNIZING DEVICE WITH PATTERN MATCHING IN SLANT PARALLELOGRAMMIC BLOCKS OF WIDTHS DEPENDENT ON CLASSIFIED REFERENCE PATTERN LENGTHS

机译:与参考基准图案长度相关的宽度倾斜倾斜的古文字学区块中具有图案匹配的图案识别装置

摘要

ABSTRACT OF THE DISCLOSURE:On pattern matching an input pattern arranged alonga first time axis to each concatenation of patterns selectedfrom reference patterns and arranged along a second time axiswith feature vectors of the input pattern and the concatenationplaced at a common frame period, a dynamic programming algorithmis used by classifying the reference patterns into classes incompliance with reference pattern lengths and by giving variableblock widths to slant parallelogrammic blocks of a common blockslope. The variable widths should be decided for the respectiveclasses. Preferably, a k-th class comprises a reference patternof a reference pattern length which is not shorter than .beta.kwand is shorter than .beta.(k + l)w, where .beta. represents the blockslope and W, a fundamental block width determined by the blockslope under a minimum of the reference pattern lengths, In thisevent, the variable block width is equal to kw. The algorithmis carried out by scaling the first time axis, not by the frameperiod, but by block numbers assigned to slant parallelogrammicblocks having the fundamental width in common and by checkingwhether or not each block number is an integral multiple of k.
机译:披露摘要:在模式匹配输入模式时所选模式的每个串联的第一时间轴从参考模式开始并沿第二时间轴排列具有输入模式和串联的特征向量放置在共同的帧周期内,采用动态规划算法通过将参考模式分类为符合参考图案长度并通过赋予变量块的宽度到普通块的倾斜平行四边形块坡。应根据各自的情况决定可变宽度类。优选地,第k类包括参考图案不小于βkw的参考图案长度的并且比β(k + 1)w短,其中β为代表方块斜率和W,由块确定的基本块宽度在最小参考图案长度下的斜率在这种情况下,可变块的宽度等于kw。算法通过缩放第一个时间轴而不是帧来执行周期,但按分配给倾斜平行四边形的块号基本宽度相同且通过检查获得的块每个块号是否为k的整数倍。

著录项

  • 公开/公告号CA1241118A

    专利类型

  • 公开/公告日1988-08-23

    原文格式PDF

  • 申请/专利权人 NEC CORPORATION;

    申请/专利号CA19850497938

  • 发明设计人 WATARI MASAO;

    申请日1985-12-18

  • 分类号G10L5/06;

  • 国家 CA

  • 入库时间 2022-08-22 06:57:03

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