High resolution disaster data clustering using Graphics Processing Units

Abstract

Near real time processing and clusters extraction from high-resolution satellite images of disaster affected area aids in monitoring and deployment of rescue activities. In this work the k-medoids clustering algorithm is analyzed for near real time applications. In general, due to the large size of satellite data, the computational time of traditional k-medoids is found to be very high. Hence in order to achieve the aim of near real time processing of such huge data we developed a parallel implementation of k-medoids (GPUPAM) and integrated with CLARA (Clustering for LARge Application), which is one of the variant of kmedoids. The implementation is performed on NVIDIA's Graphical Processing Unit (GPU).The performance improvements that were obtained is demonstrated by a GPU implementation on high resolution Haiti Earthquake QuickBird (2.4 meter resolution) data and compared with the traditional sequential implementation. The results show that the GPU implementation is found to achieve almost 96% performance improvement as compared to the sequential implementation.

Parallelization of Partitioning Around Medoids (PAM) in K-Medoids Clustering on GPU

“…Parallel k-medoids clustering can also be implemented on Graphics Processing Unit (GPU). Several GPU accelerated researchers have developed k-medoids clustering: parallel PAM implementation using CUDA [24] [25], GPU based parallel k-medoids (combined PAM-CLARA) clustering for remote sensing data [26], and GPU accelerated parallel clustering algorithms including k-means clustering, k-medoids clustering, and hierarchical clustering [27].…”

Towards extending the SWITCH platform for time-critical, cloud-based CUDA applications: Job scheduling parameters influencing performance

“…Goodman et al [20] analysed marine imaging spectroscopy data using a method derived from that of Lee et al [21,22,23]. A few papers concerned the implementation of methods for processing satellite images, namely Kurte and Durbha [24], Bhangale and Durbha [25], and Scott et al [26]. Relating to non-image processing applications, Christgau et al [27] implemented the tsunami simulation algorithm EasyWave [28] in parallel.…”