Aurora image classification method based on latent theme combining with saliency information

Abstract

The invention discloses an aurora image classification method based on a latent theme combining with saliency information, and mainly solves the problem that existing technical classification is low in accuracy and classification efficiency and narrow in application range. The method includes the implementation steps: (1) preprocessing an aurora image, extracting visual words of the preprocessed aurora image and generating a visual documentation; (2) using a spectral residual algorithm to acquire an aurora saliency map of the inputted aurora image, extracting visual words of the aurora saliency map and generating a visual document of the aurora saliency map; (3) connecting the visual documents in the step (1) and the step (2) to generate a semantic enhanced document of the aurora image, and inputting the semantic enhanced document of the aurora image to a Latent Dirichlet Allocation model to obtain saliency information latent semantic distribution characteristics SM-LDA of the aurora image; (4) inputting the SM-LDA characteristics into a support vector machine for classification so as to obtain a final classification result. By the method applicable to scene classification and target recognition, high classification accuracy is maintained, meanwhile, classification time is shortened, and classification efficiency is improved.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to an aurora image classification method, which can be used for scene classification and target recognition. Background technique [0002] The aurora is a brilliant and beautiful brilliance that is produced at night in the high altitudes near the earth's north and south poles due to the entry of charged particles from the sun into the earth's magnetic field. The occurrence of auroral phenomena often affects radio communication, long cable communication, etc. Aurora can also affect climate and biological activities. There are many forms of aurora, and different forms of aurora contain different physical meanings, so the classification of aurora images has very important scientific research value. [0003] The classification of auroral images has developed from the early manual marking by naked eye observation to the current quantitative analysis by computer. [0004] In 1...

AI Technical Summary

Problems solved by technology

In 2008, Gao Lingjun proposed a method based on the radial texture features of coronal auroras in the document "Gao Lingjun, Gabor Transform-Based Research on the Classification Algorithm of Sunside Aurora [Master's Dissertation]. Xi'an: Xidian University, 2009". Gabor transform day-side aurora classification algorithm, this method uses the Gabor filter bank to extract the local Gabor features of the aurora image, this feature can be more consistent with the texture characteristics of the coronal aurora emission, but it is not generalizable to other types of aurora images

In 2010, Wang et al. in the literature "Wang Y, Gao X, Fu R, et al. Dayside corona aurora classification based on X-gray level aura matrices. Proceedings of the ACM International Conference on Image and Video Retrieval. ACM, 2010:282 In -287", according to the characteristics of the aurora shape, a day-side aurora image classification algorithm based on X-GLAM features is proposed. This method is designed for the special texture characteristics of the corona aurora, and enhances the ability to express directional features. In terms of the influence of illumination and rotation It has strong robustness, but it performs a large number of pixel operations, which is time-consuming

However, the feature description obtained by inputting the image into the SPM model is often high-dimensional data, which increases the classification time of the image and requires more machine memory.

LDA (Latent Dirichlet Allocation) model, see the literature "Blei, D.M., Ng, A.Y., Jordan, M.I.Latent Dirichlet Allocation.Journal of Machine Learning Research.3:993-1022,2003", can effectively reduce the feature dimension of the image, But it will reduce the final image classification accuracy

Images