摘要
曲率尺度空间图像轮廓线的最大值可以用来表达物体边界并与2维平面图像对应。由存储数据库随机调用相应标准图像可以识别未知物体。采用有限少数图像并修复形变图像轮廓,尝试有条件地识别三维物体。
the curvature scale space image contours could be used as a representation of a object boundary which can be considered as a two dimension image shape, hence we could identify an unknown object from an image taken from a random view in a standard image database. By limited small images and retrieval of the contour an object shape can be descript at certain constraint conditions.
出处
《武汉科技学院学报》
2006年第6期13-15,共3页
Journal of Wuhan Institute of Science and Technology
基金
湖北省教育厅重点项目
教育部重点项目科技项目(.项目编号:2000Z04007)
湖北工业大学科研启动基金资助
关键词
图像识别
曲率尺度空间
形变
recognition
curvature scale space
shape deformation
