Edge Camera based Dynamic Lighting Control System for Smart Streetlights

In this paper, we propose a novel dynamic lighting control system for smart streetlights using an edge camera which detects objects for itself using a light weight deep neural network at local. The proposed system consists of an edge camera, a light control unit, and a control server. The proposed system dynamically adjusts the brightness of a dimming module based on event messages from the edge camera. The proposed system maintains low brightness when there are no pedestrians and switches to high brightness when pedestrians appear. And when a pedestrian leaves the service radius, it returns to low brightness. We designed the protocol between the edge camera and the light control unit, and developed the pedestrian detector using a lightweight deep neural network, which provides real-time processing and the high detection accuracy.

[1]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Ren Li,et al.  Imbalanced SVM‐Based Anomaly Detection Algorithm for Imbalanced Training Datasets , 2017 .

[3]  Ali Farhadi,et al.  YOLO9000: Better, Faster, Stronger , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Bo Chen,et al.  MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications , 2017, ArXiv.

[5]  Forrest N. Iandola,et al.  SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <1MB model size , 2016, ArXiv.

[6]  S. Karthik,et al.  Intelligent Street Lights , 2015 .

[7]  Ali Farhadi,et al.  YOLOv3: An Incremental Improvement , 2018, ArXiv.

[8]  Yongjin Kwon,et al.  Vision‐based garbage dumping action detection for real‐world surveillance platform , 2019, ETRI Journal.

[9]  Wei Liu,et al.  SSD: Single Shot MultiBox Detector , 2015, ECCV.

[10]  Mark Sandler,et al.  MobileNetV2: Inverted Residuals and Linear Bottlenecks , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[11]  Minji Kim,et al.  Robust appearance feature learning using pixel‐wise discrimination for visual tracking , 2019 .

[12]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[13]  Charles X. Ling,et al.  Pelee: A Real-Time Object Detection System on Mobile Devices , 2018, NeurIPS.