Aerospace Technic and Technology

The subject matter of the article is the methods of automatic clustering of remote sensing data under conditions of a priori uncertainty regarding the number of observed object classes and the statistical characteristics of the signatures of classes. The aim is to develop a method for approximating multimodal empirical distributions of observational data to construct decision rules for pixel-by-pixel statistical classification procedures, as well as to investigate the effectiveness of this method for automatically classifying objects on synthesized and real images. The tasks to be solved are: to develop and implement a procedure for splitting a mixture of basic distributions, while ensuring the following requirements: the absence of a preliminary data analysis stage in order to select optimal initial approximations; a good convergence of the method and the ability to automatically refine the list of classes by combining indistinguishable or poorly distinguishable components of the mixture into a single cluster; to synthesize test images with a specified number of objects and known data distributions for each object; to evaluate the effectiveness of the developed method for automatic classification by the criterion of the probability of correct recognition; to evaluate the results of automatic clustering of real images. The methods used are methods of stochastic simulation, methods of approximation of empirical distributions, statistical methods of recognition, methods of probability theory and mathematical statistics. The following results have been obtained. A method for automatic splitting of a mixture of Gaussian distributions to construct decision thresholds according to the maximal a posteriori probability criterion was proposed. The results of the automatic forming the list of classes and their probabilistic descriptions, as well as the results of the clustering both test images and satellite ones are given. It is shown that the developed method is quite effective and can be used to determine the number of objects’ classes as well as their stochastic characteristics’ mathematical description for pattern recognition tasks and cluster analysis. Conclusions. The scientific novelty of the results obtained is that the proposed approach makes it possible directly during the “unsupervised” training procedure to evaluate the distinguishability of classes and exclude indistinguishable objects from the list of classes.

pattern recognition; clustering; approximation; a mixture of basis functions; estimation of the mixture parameters; the probability of correct recognition

Fukunaga, K. Vvedenie v statisticheskuyu teoriyu raspoznavaniya obrazov [Introduction to statistical theory of pattern recognition]. Moscow, Nauka Publ., 1979. 368 p.

Aivazyan, S. A., Bukhshtaber, V. M., Enyu-kov, I. S., Meshalkin, L. D. Prikladnaya statistika: klassifikatsiya i snizhenie razmernosti [Applied Statistics: Classification and Dimension Reduction]. Moscow, Finansy i statistika Publ., 1989. 608 p.

He, W., Yu, R., Zheng, Y. and Jiang, T. Image Denoising Using Asymmetric Gaussian Mixture Models. Internat. Symposium in Sensing and Instrumentation in IoT Era (ISSI), Shanghai, 2018, pp. 1-4. doi: 10.1109/ISSI.2018.8538279.

Sun, J., Zhao, Y., Wang, S. Image Compression Using GMM Model Optimization. IEEE Proc. Internat. Conf. on Acoustics, Speech and Signal Processing (ICASSP), Brighton, United Kingdom, 2019, pp. 1797-1801. doi: 10.1109/ICASSP.2019.8683784.

Yang, G., Li, H., Yang, W., Fu, K., Celik, T. and Emery, W. J. Variational Bayesian Change Detection of Remote Sensing Images Based on Spatially Variant Gaussian Mixture Model and Separability Criterion. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2019, vol. 12, no. 3, pp. 849-861. doi: 10.1109/JSTARS.2019.2896233.

Bouhlel, N., Méric, S. Unsupervised Segmentation of Multilook Polarimetric Synthetic Aperture Radar Images. IEEE Transactions on Geoscience and Remote Sensing, 2019, pp. 1-15. doi: 10.1109/TGRS. 2019.2904401.

Shahin, I., Nassif, A. B. and Hamsa, S. Emotion Recognition Using Hybrid Gaussian Mixture Model and Deep Neural Network. IEEE Access, 2019, vol. 7, pp. 26777-26787. doi: 10.1109/ACCESS.2019. 2901352.

Hollaus, F., Diem, M. and Sablatnig, R. MultiSpectral Image Binarization using GMMs. 16th Internat. Conf. on Frontiers in Handwriting Recognition (ICFHR), Niagara Falls, NY, 2018, pp. 570-575. doi: 10.1109/ICFHR-2018.2018.00105.

Popov, A. V., Pogrebnyak, O., Brashevan, A. N. Image clustering algorithm using polynormal distribution. Proc. of SPIE Mathematical Methods in Pattern and Image Analysis, 2005, vol. 5916, pp. 1-9.

Vasil'eva, I. K. Iteratsionnyi metod otsenki parametrov smesi funktsii Gaussa v zadachakh opisaniya dannykh nablyudenii [An iterative method for estimating the parameters of a mixture of Gauss functions in problems of describing observational data]. Radioelektronni i komp’juterni systemy – Radioelectronic and computer systems, 2015, no. 3 (73), pp. 70-76. (In Russian).

Vasil'eva, I., Popov, А. Multicomponent Model of Objects Attributive Signatures on Color Pictures. Proc. Internat. Scientific-Practical Conf. on Problems of Infocommunications. Science and Technology, Kharkiv, Ukraine, 9-12 Oct. 2018, pp. 281-284. doi: 10.1109/INFOCOMMST.2018.8632110.