Development of multiresolution hierarchical trees to non-cooperative target recognition

Abstract

To cite this version:Christian Brousseau. Development of multiresolution hierarchical trees to non-cooperative target recognition. MTA Review -Military Technical Academy Publishing House, 2012, XXII (2)

Development of a tree structured hierarchical wavelet representation of synthetic database to NCTR

“…Previous results [5] has shown that better performances are obtained with a hard clustering algorithm, like K-means, in NCTR applications. Thus, in the next sections, results obtained with the K-means hard partitioning method are only presented.…”

“…Fig.4 presents an example of range profile and its wavelet decomposition computed in four levels. Another work [5] has shown that there is no statistically significant difference in performance of the classifier when different wavelets are chosen. Thus, in the next sections, results obtained with the Haar wavelet are only presented.…”

“…Thus, multiresolution hierarchical trees are a solution to compress high resolution data of radar targets. But, other methods exist like the use of wavelet decomposition or the unsupervised data clustering [5]. It must be interesting to compare these techniques as a function of probability of false classification and computational time of search.…”

“…Clustering deals with finding a structure in a collection of unlabeled data. But each of them has its own limitations [5]. Application of wavelets representation to NCTR application gives a low decrease of recognition search time but with a low degradation of probability of false identification.…”

Estimation of efficiency of a tree structured hierarchical wavelet representation of synthetic database applied to non-cooperative target recognition

“…A previous work [15] has shown that there is no statistically significant difference in performance of the classifier when different wavelets are chosen. Thus, in the next sections, only results obtained with the Haar wavelet are presented.…”