ABSTRACT:
There has been a tremendous development in the amount of visual data available on the Internet in recent years. One kind the of visual data particular interest is produced by network cameras giving real-time views. Millions of network cameras around the world continuously stream information to viewers connect with the Internet. This information might be utilized by a wide variety of utilization, for example, improving public security, urban planning, emergency response, and traffic management which are computationally intensive.
Cloud computing can be a preferred solution for meeting the resource necessities for analyzing this information. There are many choices while choosing cloud instances and inefficient provisioning of cloud resources may become to be expensive in pay-per-utilize distributed computing. This project displays a technique to choose cloud occasions keeping in mind the end goal to meet the execution prerequisites for visual information analysis at the lower cost. We define the issue of managing cloud resources as a Variable Size Bin Packing Problem and utilize a heuristic solution. Experiments utilizing Amazon EC2 approve the model and show that the proposed solution can decrease the cost by up to while meeting the performance requirements.
