SCEM: Smart & effective crowd management with a novel scheme of big data analytics
Shakti Awaghad · 2016
The proposed paper presents a novel scheme that can perform a precise extraction of knowledge from the complex and massive streaming of live data of the scene from the crowded place. The prime contribution of the proposed system is to perform enough processing over the raw and unstructured distributed data from multiple locations so that processing over distributed storage and mining can be done with lesser processing time and higher degree of accuracy. An experimental research methodology has been adopted to capture signal using Logitech HD C920 and processed over Intel Xeon E5540 processors with 2 GPbs connectivity. The raw data is subjected to pre-processing, segmentation, scene profiling, in order to get convolved data that are stored in distributive manner using Hadoop and mined using MapReduce. The comparative study outcome shows lesser processing time and higher accuracy as compared to existing relevant analytics.