A Streaming Cloud Platform for Real-Time Video Processing on Embedded Devices

Real-time intelligent video processing on embedded devices with low power consumption can be useful for applications like drone surveillance, smart cars, and more. However, the limited resources of is a challenging issue effective computing. Most existing work this topic focuses single device based solutions, without use cloud computing mechanisms parallel to boost performance. In paper, we propose platform real-time devices. Eight NVIDIA Jetson TX1 three TX2 GPUs are used construct streaming (SECP), which Apache Storm deployed as environment deep learning algorithms (Convolutional Neural Networks - CNNs) process streams. Additionally, self-managing services designed ensure that run smoothly stably, in form metric sensor, bottleneck detector scheduler. This evaluated terms speed, consumption, network throughput by running various object detection. The results show proposed while meeting high scalability fault tolerance required processing.