| Summary: | 碩士 === 國立臺北科技大學 === 電機工程研究所 === 104 === Recently, the information of ground surface has been recode with Hyperspectral device massively and recognition the ground material by analyzing the spectral data. The k-Nearest Neighbor (k-NN) and Nearest Feature Line(NFL) algorithm is widely used in classify, the main idea of k-NN algorithm is that find the k nearest neighbor and voting by their class ID, and the main idea of NFL algorithm is to find the nearest distance making by two train sample and voting by their class ID. However, the overlapping of different training sample groups will cause false classification. To overcome this problem, we trying to use Nearest Feature Space(NFS) algorithm to keep the structure of training samples and calculate the nearest distance between test samples and the feature space of training samples.
Since, the NFS algorithm is mainly used in 3D space to do classify, but the algorithm can also be used in higher than 3D space, so we provided “Nearest Feature HyperSpace Algorithm” to do classify, and compared with the other classifiers, discussed which classifier can get better recognition results.
|
|---|
