No abstract

No abstract

No abstract

RESUMEUn nouveau modele pour la detection automatique des bateaux cibles dans les images SAR est developpe dans ce papier. Le modele utilise une methode statistique combinee avec les techniques de traitement d'images. Le reseau de neurones probabilistiques (PNN) est un modele ires efficace pour la classification des donnees. Il est base sur l' approche non-parametrique pour l' estimation des densites de probabilites. Ce modele statistique donne toujours de bons resultats quand il est utilise dans l' estimation des densites des histogrammes d'images a niveaux de gris. La technique CFAR basee sur Ie PNN est utilisee dans ce papier pour ameliorer la performance de la detection des bateaux cibles dans les images SAR. Ce papier presente nos recents travaux realises pour l' amelioration du modele de detection. SUMMARY A new automatic detection modelfor ship target in SAR imagery is being developed by using a statistical method and image processing technique. Probabilistic neural networks (PNN) are an efficient model for data classification based on a non-parametric approach to estimate probability density function. This statistical model always gives a good fit for gray-level image histograms ofSAR images. The CFAR technique based on the PNN-model is used to improve ship target detection performance. This paper presents current progress made on the detection model.

Common methods for matching multivariate time series such as the Euclid method and PCA method have difficulties in taking advantage of the global shape of time series. The Euclid method is not robust, while the PCA method is not suitable to deal with the small-scale multivariate time series. This paper proposes a pattern matching method based on point distribution for multivariate time series, which is able to characterize the shape of series. Local important points of a multivariate time series and their distribution are used to construct the pattern vector. To match pattern of multivariate time series, the Euclid norm is used to measure the similarity between the pattern vectors. The global shape characteristic is used in the method to match patterns of series. The results of experiments show that it is easy to characterize the shape of multivariate time series with this method, with which various scales can be dealt with in series data.