New Object Detection, Tracking, and Recognition Approaches for Video Surveillance Using Hadoop
Mohan Shinde, ShashankSupe, Sushant Shelar, Mayur Shinde · International journal of advance research and innovative ideas in education · 2016
Object detection and tracking are two fundamental tasks in multicamera surveillance. The most important technique of this multicamera related technique is to track and analyze objects within the images. The core technology of multicamera analysis is used in detecting, analyzing, and tracking the object’s motion. In addition, when the light’s color or direction changes, it is difficult to trace the object.Firstly use the block based algorithm for detecting the change scene in video if the scene is change is detected then video is stored on the server for further analysis. Once the video was stored on the server. Stored videos are dived in to chunks and send to different nodes for analysis using map reduce technology of Hadoop. for detecting object, we apply algorithms like SSIM index, Histogram matching Using Hadoop we minimize the analysis time Finally draw the graphs in which show the no of objects to be detected and time to be required for analysisand stored analysis result into database for security purpose.