发明名称 IMAGE CONTOUR EXTRACTING METHOD
摘要 PURPOSE:To exactly demarcate the contour of a symmetrical object even when a contour can not be extracted due to a luminance change by calculating and synthesizing first edge intensity data and second edge intensity data obtained by the difference processing based on coordinate data. CONSTITUTION:The image data of the symmetrical object 3 obtained by a three-dimensional image input device 1 is composed of the luminance data of the surface of the symmetrical object 3 and distance data expressed by xyz coordinate data. In a first difference processing circuit 10, a difference processing is performed for luminance data and first edge intensity data is calculated and derived. In a second difference processing circuit 11, the difference processing is performed for distance data, second edge intensity data is calculated and derived. A binary processing is performed for each data by proper thresholds by using binary circuits 12 and 13. Thus, the contour data of the symmetrical object 3 is extracted from each of luminance data and distance data. By determining the difference of the OR and AND of the both extracted contour data by using a synthetic processing circuit 14, the contour data of the symmetrical object 3 is determined.
申请公布号 JPH0793562(A) 申请公布日期 1995.04.07
申请号 JP19930234638 申请日期 1993.09.21
申请人 KUBOTA CORP 发明人 TANAKA YOSHINORI;HORIKIRI KOJI
分类号 H04N7/18;G06T7/60;G06T9/20 主分类号 H04N7/18
代理机构 代理人
主权项
地址