Optimization of Video Processing for Real-time Image Recognition

EhEye Inc.是一家加拿大初创公司，该公司开发了利用现有的摄像头基础设施自动检测武器、骚乱和感兴趣的事件的技术，用于安全目的。该技术允许实时检测不规则行为，以便位置所有者/经理/安全部门可以自动检测与他们的需求最相关的事件。这项技术已经部署在各种不同的环境中，包括大规模的人群和集会。随着公司继续扩大规模，对其产品的需求不断增加，优化其系统和算法的性能以处理许多相机领域将变得至关重要。EhEye Inc.已聘请我们调查和总结与图像识别系统优化相关的文献中的最新技术，并就优化其算法和系统的性能的适当技术提供一套建议。

EhEye Inc. is a Canadian start-up company that has developed techniques to automate the detection ofweapons, disturbances, and incidents of interest using existing camera infrastructures for security purposes.The technology allows for the real-time detection of irregular behavior so that locationowners/managers/security can automate the detection of incidents that are most relevant to their needs. Thetechnology has been deployed in a variety of settings, including large crowds and gatherings. As the companycontinues to scale, and demand for their product increases, optimizing the performance of their systems andalgorithms to handle many camera fields will become paramount. EhEye Inc. has engaged us to survey andsummarize the state-of-the-art in the literature as it relates to the optimization of image-recognition systems,and to provide a set of recommendations on appropriate techniques to optimize the performance of theiralgorithms and systems.