| 摘要 |
As shown in FIG.7, a visual device(2) detects position, size and form of at least one object in a digital image (111), by using a geometrical analysis means (37). In addition, the visual device (2) detects position, size, inclination, pattern, texture and color of a segmented object area (143) in a digital image (111) corresponding to at least one object area (141) separated from a background area by a figure/ground separation means (16). Therefore, the visual device (2) can make an environmental map for the object by using an environmental understanding means (31). As shown in FIG.8, moreover, a visual device (2) can take a digital image (111) which shows the object whose size is suitable at its center, by controlling a moving camera (10). <IMAGE> |
