Implementation of Fire and Smoke Detection using DeepStream and Edge Computing Approachs

摘要

The deep neural network has made a significant result in many fields. The computing ability has become an efficient tool to solve a complex problem such as face detection, automatic driving. However, deep learning models rely on high-performance servers equipped with powerful computing processors and large storage. In this research, we try to implement fire detection on the Jetson NX Xavier as Edge Devices. Nvidia Jetson has feature CUDA GPU where can be programmed to accelerate complex machine learning. Deepstream pipeline and YOLOv3 have successfully implemented in the Jetson NX Xavier to detect fire and smoke.