Abstract
Video data have become critical source of information that can be used in many video surveillance applications. Processing large video data can benefit from a technology such as Apache Hadoop that provides a transformative solution for distributed storing and processing of large data. The goal of this paper is to discuss how to improve the performance of Change Detection while processing large video data on Hadoop clusters. The proposed approach uses two sampling techniques along with MapReduce programing model to improve the execution time of change detection. The experiments have been conducted in the Cloud (Amazon EC2). The experimental results exhibit that the proposed approach improves the execution time of MapReduce-based change detection.