| 摘要 |
The present invention is provided with: a three-dimensional object detection means (32, 33, 34) that generates difference waveform information by counting the number of pixels indicating a difference equal to or greater than a difference threshold value in a difference image in which overhead view images taken at different times are aligned, and detects a three-dimensional object on the basis of the difference waveform information; a reference image area setting means (35) that sets, as a reference image area, an area which is different from a detection area and from which the extraction of contour information of an object is predicted; a contour information detection means (35) that extracts the contour information of the object; a reference sharpness calculation means (35) that calculates the reference sharpness on the basis of the contour information detected in the reference image area; and a detection-target sharpness calculation means (35) that calculates the detection-target sharpness on the basis of the contour information detected in the detection area. The three-dimensional object detection means (32, 33, 34) calculates the final sharpness on the basis of the reference sharpness and the detection-target sharpness, and sets the difference threshold value on the basis of the calculated final sharpness. |
