Cloud Resource Management for Network Cameras

Most of the network cameras flow dynamic and adequate visual data from several locations such as traffic, check gate/tollgate, natural scenes, etc. These data from high resolution network cameras can be used in several applications that help to find the needed significant resources. The cloud vendors provide the required resources as per the hourly cost which is not always so affordable for several applications using cameras and instances. This problem can be solved using bin packing and heuristic algorithm. Thus, the allocation and deallocation of instances are maintained by cloud resource manager, by which the cost is controlled accordingly. The performance of this framework is considered based on multiple analyses of programs, moving object detection, feature tracking, and large number of images.