| 摘要 |
<p>PURPOSE:To detect a moving body from the 2nd original image without forming a background image based upon the small volume of information by extracting a part of a new image corresponding to a new outline part out of the outline of a new image. CONSTITUTION:The 1st inputted image is regarded as an old image to store the outline 30A of the old image, the 2nd inputted image is regarded as a new image 3A and a new outline part 43 out of the outline 29A of the new image 3A is discriminated from the outline 30A of the old image. When a part 42 of the new image 3A corresponding to the part 43 is extracted, the extracted part 42 corresponds to a changed part, i.e. a moving body, against the old image. Consequently, the moving body can be detected immediately after inputting the original image without forming its background image and an incomplete moving body also can be accurately detected.</p> |
