DefinePK

Abstract

The proposed research work focuses on video analytics which is been emerged as the fastest growing research area in the schematic sketch of Big data and its applications in cloud. The exploration of video is been well marginalized and organized by Hadoop Infrastructure, whose propagation is discussed. Large volume of unstructured data is been produced in day-to-day activities of the people, not exempting various organizations that are recorded and maintained eventually for research and other purposes. Traditional database architectures were not able to handle the generation, storage and access of this huge amount of unstructured (video) data. This proliferation led to the collaboration of video-analytics with a conservative database theory called Big Data. There are several structured database that are developed to master the needs for accessing and maintaining the videos. Despite, these developments, the retrieval and processing of the particular environment detail, specifically a featured human, or an object from the videos takes huge amount of time which is not too effective. So, this shortcoming is been forwarded to the Hadoop Distributed Video File System (HDVFS) whose Map-Reduce Framework process the recognition of stipulated image/object and their behavior from the unstructured video in the cloud repository which stores the intensive-sized files like Amazon S3 storage bucket.