Parallel implementation of multiple kernel self-organizing maps for spectral unmixing
- PDF / 2,950,067 Bytes
- 18 Pages / 595.276 x 790.866 pts Page_size
- 24 Downloads / 148 Views
ORIGINAL RESEARCH PAPER
Parallel implementation of multiple kernel self‑organizing maps for spectral unmixing Ghada M. Fathy1 · Hanan A. Hassan1 · Shaheera Rahwan1 · Walaa M. Sheta1 Received: 16 October 2018 / Accepted: 4 May 2019 © Springer-Verlag GmbH Germany, part of Springer Nature 2019
Abstract Spectral unmixing algorithms are commonly used in processing of hyperspectral images to identify the elemental components, called end-members, and their corresponding information in each pixel of the image. However, these algorithms are computationally intensive and can become a bottleneck for remote sensing hyperspectral image processing, especially in large aerial imagery processing centers. This paper, explores the use of massive parallel processing graphical processing unit to speed up the multi kernel self-organizing map (MKSOM) unmixing algorithm. MKSOM is based on artificial neural networks, which makes it suitable to be efficiently parallelized. Two real benchmark hyperspectral images; AVIRIS Cuprite and Brullus are used to evaluate the performance of the parallel algorithm. The experimental results show that the proposed implementation is appropriated for real-time hyperspectral remote sensing applications due to a very small worst case parallel execution time (0.83 s when the number of classes is less than 9) which makes it feasible to be integrated as on-board processing on any Hyperspectral remote sensors. Our parallel technique achieved a significant speedup compared with a multi-threaded CPU implementation applied on the same hyperspectral image. The results showed a speedup of 93.46 × for SOM size of 256 and trained for 100 epochs on medium-sized HSI such as AVIRIS Cuprite. Keywords Spectral unmixing · Hyperspectral image · GPU · Self-organization map · Remote sensing applications
1 Introduction Hyperspectral imaging technologies are considered nowadays as one of the most powerful tools for acquiring surface information in many different fields, such as environmental mapping, risk prevention, urban planning, pollution monitoring and mining exploration [22]. These systems are able to provide images in which single pixels have information from across the electromagnetic spectrum of the scene under observation. Hyperspectral images can be obtained by satellite or airborne sensors, which collect hundreds * Ghada M. Fathy [email protected] Hanan A. Hassan [email protected] Shaheera Rahwan [email protected] Walaa M. Sheta [email protected] 1
Informatics Research Institute, SRTA City, Alexandria, Egypt
or even thousands of spectral bands. Spectral unmixing is an important task for remotely sensed hyperspectral data exploitation and is used to solve various problems such as mixed pixels problem [7]. Spectral unmixing [5] solve this problem by identifying pure spectral components and their abundance fractions in each pixel. This is achieved by different approaches such as linear mixture model (LMM) and nonlinear mixture models (NLMM). The linear model has practical advantages such a
Data Loading...
