2DPCANet: Dayside Aurora Classification Based on Deep Learning
The mysterious and beautiful aurora represents various physical meaning, thus the classification of aurora images have significant scientific value to human beings. Principal component analysis network (PCANet) has achieved good results in classification.
- PDF / 1,062,479 Bytes
- 12 Pages / 439.37 x 666.14 pts Page_size
- 33 Downloads / 226 Views
School of Electronic Engineering, Xidian University, Xi'an 710071, China [email protected], [email protected] 2 State Key Laboratory of Remote Sensing Science, Beijing 100101, China
Abstract. The mysterious and beautiful aurora represents various physical meaning, thus the classification of aurora images have significant scientific value to human beings. Principal component analysis network (PCANet) has achieved good results in classification. But when using PCANet to extract the image features, it transform original image into a vector, so that the structure information of the image are missing. Compared with PCA, 2DPCA is based on 2D image matrices rather than 1D vectors so that 2DPCA can use the structure information of original image more efficiently and reduce the computational complexity. Based on PCANet, we develop a classification method of aurora images, 2-dimension PCANet (2DPCANet). To evaluate 2DPCANet performance, a series of experiments were performed on two different aurora databases. The classification rate across all experiments was higher using 2DPCANet than PCANet. The experiment results also indicated that the classification time is shorter using 2DPCANet than PCANet. Keywords: Aurora image · Deep learning · Principle component analysis · PCANet · 2DPCANet
1
Introduction
Aurora is the magnificent light when the solar wind travels the magnetosphere of high altitude areas near the north and south poles of the earth. The aurora is not only related to the earth's atmosphere and geomagnetic field, but also related to the solar eruption of high-speed charged particles. When charged particles are emitted by the sun toward the earth into the scope of the earth magnetic field, they travel along the earth's magnetic field lines into the upper atmosphere near the north and south poles under the influence of the magnetic field, and then inspire visible light after proton collisions, and finally become a high-profile, we call it aurora. Aurora phenomenon is not only a simple optical phenomenon, but also an important way for understanding the atmospheric physics. Different forms of the aurora imply different physical meanings. Therefore, the highly efficient classification of aurora images has very important value in scientific research. From 1964 till now, Akasofu [1], Hongqiao Hu [2], and the Chinese polar research center [3] have divided the aurora into different types. For a long time, the aurora was divided into arc aurora and corona aurora. The corona aurora was further divided © Springer-Verlag Berlin Heidelberg 2015 H. Zha et al. (Eds.): CCCV 2015, Part II, CCIS 547, pp. 323–334, 2015. DOI: 10.1007/978-3-662-48570-5_32
324
Z. Jia et al.
into drapery corona, radial corona and hot-spot corona. In 2015, on the basis of arc aurora and corona aurora, the Chinese polar research center considers that corona aurora contains two types aurora: drapery corona and radial corona. Hot-spot corona aurora is regarded as anther aurora. So now the most significance aurora types are arc aurora, drapery cor
Data Loading...
