|
KNN ALGORITHM IN A FRAMEWORK OF SCALE-SPACE THEORY FOR RETINAL IMAGE ANALYSIS |
|---|---|
| รหัสดีโอไอ | |
| Creator | 1. Krit Inthajak 2. Cattleya Duanggate 3. Bunyarit Uyyanovara 4. Stanislav Makhanov |
| Title | KNN ALGORITHM IN A FRAMEWORK OF SCALE-SPACE THEORY FOR RETINAL IMAGE ANALYSIS |
| Publisher | Suaranaree University of Technology |
| Publication Year | 2557 |
| Journal Title | Suranaree Journal of Science and Technology |
| Journal Vol. | 21 |
| Journal No. | 2 |
| Page no. | 87-96 |
| Keyword | Feature stability, K-nearest neighbor, object detection, scale-space |
| ISSN | 0858-849X |
| Abstract | This topic presents a framework in the uses of the K-nearest neighbor algorithm in evaluating an object detection method of scale-space theory with feature stability. A scale-space tree is constructed based on the blobs that were created from a series of images after the blurring process. Features and spatial information provide the role within the scale-space tree construction. After the process of blob extraction, users determine each type of the blob that was detected within the image by distinguishing classes to create ground truth image data. Within the same process, the KNN algorithm is applied to distinguish classes of the image |
