Independent ingredient analysis global search method for implementing high spectrum terrain classification

Abstract

The invention relates to independent component analysis global search method of realizing the high spectrum fine classification under no prior knowledge situation, the method including: reading in the high spectral data, the establishment of independent component analysis model based on the kurtosis, the center of the data, the ball of data, the iterative solution based on quantum genetic algorithm, independent component compositor, two value of the image, the feature classification. The invention method can established on the circumstance of no data background model, using the self high statistical data to achieve the feature fine unsupervised classification of the high spectral data; at the same time, avoided to plunge in the local best solution problem in the independent component analysis solution process, and compared with the traditional genetic algorithm, the invention used quantum genetic algorithm has less number of iterations, fast convergence, high search efficiency and the strong overall search capability and so on features.

Description

technical field [0001] The invention relates to the field of remote sensing information processing, in particular to a global search method based on independent component analysis applied to hyperspectral data feature extraction and fine classification of ground objects, which can effectively improve the classification accuracy of ground objects. Background technique [0002] The hyperspectral imager is a new type of remote sensing payload. Its spectrum is compact and continuous, and it can simultaneously record the spectral and spatial information characteristics of the same ground object. Therefore, hyperspectral remote sensing provides a powerful detection method for the fine classification of ground objects. At present, the methods of ground object classification using hyperspectral image data are mainly divided into two categories: supervised and unsupervised classification methods. Supervised classification methods can obtain better performance under the assumption tha...

AI Technical Summary

Problems solved by technology

[0005] In view of the problems that the ICA algorithm is easy to fall into the local optimal solution, the ICA based on the neural learning algorithm is affected by the activation function of neurons, and the fine classification algorithm of the hyperspectral supervised object is difficult to obtain prior knowledge, the present invention proposes the realization of hyperspectral unsupervised The ICA method based on the improved quantum genetic algorithm for fine classification of ground features overcomes the difficulty of obtaining prior knowledge in supervised classification methods, effectively improves the global search ability of ICA and the accuracy of hyperspectral ground feature classification

Images