| 摘要 |
PROBLEM TO BE SOLVED: To increase the accuracy of duplicate detecting algorithm. SOLUTION: At least two images which are picked up at possibly determined time are prepared, the trends of image components of each image are calculated to determine the pickup time of the image, and the trends and pickup time of the image components are evaluated so as to judge whether or not those two images are duplicate images, thereby detecting duplicate images. The images are divided into blocks and the trends of the image components are computed from the histograms of the respective blocks. Then when the trends and pickup time of the image components are evaluated, a histogram crossing metric is used to compare 1 block of one image with the corresponding block of the other image, and the difference in pickup time is sued to judge whether or not the two images are duplicate images.
|
