发明名称 IMAGE SIGNAL ENCODING METHOD
摘要 PURPOSE:To reduce the amount of codes in the case of encoding the information of feature points by extracting only the desired feature point from an input image signal. CONSTITUTION:A feature point signal S3 of '0' or '1' provided by a twodimensional change point detecting circuit 4 is stored in a frame buffer 42 of a mask image processing circuit 41. As a result, the frame buffer 42 stores a binary image composed of the feature point and a point which is not the feature point. The mask processing circuit 41 extracts a present processing point and points around it from the binary image in the frame buffer 42 and generates a ROM address designate signal S47 by arranging those points in the prescribed order. Based on the ROM address designate signal S47 and the numbers of feature points and the points which are not the feature points contained in the ROM address designate signal S47, those points are inputted to a deciding ROM 51 where '0' or '1' is stored in advance. When the point is the feature point to be defined as an encoding object, '1' is outputted from the deciding ROM 51 as a feature point signal S30 but when it is the feature point not to be defined as the encoding object, '0' is outputted.
申请公布号 JPH087110(A) 申请公布日期 1996.01.12
申请号 JP19940158096 申请日期 1994.06.16
申请人 SONY CORP 发明人 MURAYAMA ATSUSHI
分类号 H04N19/102;G06T9/00;H03M7/30;H04N1/41;H04N7/24;H04N19/00;H04N19/134;H04N19/136;H04N19/196;H04N19/20;H04N19/21;H04N19/423;H04N19/59 主分类号 H04N19/102
代理机构 代理人
主权项
地址
您可能感兴趣的专利