Image Duplication and Rotation Algorithms for Storage Utilization
In this era, there is high end storage with high storage capacity. However, due to ethic issues, this development has rapidly increased the cost of hardware every year. In order to prevent the rapid increase of the hardware cost, pattern detection methods are introduced to optimize the storage utilization by detecting image duplication and inappropriate image. Pattern detection is usually applied before pattern recognition. The accuracy of pattern detection can give impact to pattern recognition. Therefore, it plays an important role on a digital image. This paper is based on scale invariant feature transform (SIFT) for image near-duplication detection and V-J face detection. V-J method refers to Viola Jones’s method. Both methods have been successfully applied to the real world problem. Since the V-J face detection method is not trained with rotated features, it has limitiation to rotation invariant, whereas the SIFT method detects many feature keypoints, affecting the speed performance. In order to overcome the issues above, this paper proposed a new method by enhancing the SIFT with speed performance and V-J face detection with high rotation invariant. The comparison result of the experiment shows that the proposed solutions produce better performance
SIFT; V-J Face Detection; Pattern Detection
- There are currently no refbacks.
This work is licensed under a Creative Commons Attribution 3.0 License.