(2D)2k-NNDA: Two-directional two-dimensional k-nearest neighbour discriminant analysis for target recognition

An image feature extraction technique, two-directional two-dimensional k-nearest neighbour discriminant analysis ((2D)
²
k-NNDA), is presented from the viewpoint of the k-nearest neighbour (k-NN) classification, which is an extension of 2DNNDA based the idea of the nearest neighbour (1-NN) classification. Similar to 2DNNDA, (2D)
²
k-NNDA makes use of the matrix representation of images and does not assume the class densities belong to any particular parametric family. (2D)
²
k-NNDA is applied to target recognition and the results demonstrate that (2D)
²
k-NNDA achieves at least the same or even higher recognition accuracy than the existing 2D subspace methods.