摘要 |
PROBLEM TO BE SOLVED: To easily obtain the distance to an existent subject, specially, in a city scene and accurately compose a scene image by using this value by making good use of subject data representing the shape of the subject. SOLUTION: An input image processor 17 parameterizes an input image with its gaze and stores a storage device 19 with the image and its gaze, and gaze information. A gaze indication device 20 specifies a new gaze position and an optical-axis direction for image composition and a gaze tracking device 21, on the other hand, finds the intersection of the gaze and shape data stored in a shape storage device 22. Further, a retrieval device 24 retrieves a partial image of the input image by using the gaze information and intersection information and a gaze determination device 18 determines a gaze vector for the partial image of the composite image. Here, a composite image generating device 23 composes an image having the gaze and optical-axis direction inputted from the indication device 20. Consequently, the distances to the subject are obtained at the respective parts of the image and the composite image accurately representing the subject can be obtained. |