Construction Safety Surveillance Using Machine Learning

Safety has always been a matter of concern in all industrial activities, especially construction. Hard hats or safety helmets act as the first line of protection against serious head injuries. Some workers don’t quite follow the instructions and signs that require them to follow safety measures. And in case of any accidents, the workers will not be eligible for insurance, if they were not wearing the safety equipments. This applied research paper mainly focuses to detect persons with and without a helmet in the construction site. The accuracy and performance of Neural Networks were tested and compared with other hand crafted features like Haar and LBP classifiers, Histogram of Oriented Gradients and Sequential Classifiers. These hand crafted features gave more false detections than neural nets, when tested in real conditions. Different Neural Networks were tested on Edge Devices such as Nvidia Jetson TX2 and Jetson Nano, for commercial deployment. Compared to the other neural networks, the SSD MobileNet model showed better performance without considerable drop in accuracy, when tested on edge devices in real-time. This makes it a preferable solution for this application-oriented problem.

[1]  Xin Pan,et al.  YouTube-BoundingBoxes: A Large High-Precision Human-Annotated Data Set for Object Detection in Video , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[3]  Pietro Perona,et al.  Microsoft COCO: Common Objects in Context , 2014, ECCV.

[4]  Xindong Wu,et al.  Object Detection With Deep Learning: A Review , 2018, IEEE Transactions on Neural Networks and Learning Systems.

[5]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[6]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[7]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

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

[9]  Tara N. Sainath,et al.  Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research Groups , 2012, IEEE Signal Processing Magazine.

[10]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[11]  Yuan Yu,et al.  TensorFlow: A system for large-scale machine learning , 2016, OSDI.

[12]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[13]  Yee Whye Teh,et al.  A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[14]  Sharath Pankanti,et al.  Smart Video Surveillance , 2005 .

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

[16]  N. Rekha,et al.  Face Detection in Real Time Based on HOG , 2014 .

[17]  アインシュタイン，リチャード，ディー,et al.  Video surveillance system , 2010 .

[18]  Sergio Guadarrama,et al.  Speed/Accuracy Trade-Offs for Modern Convolutional Object Detectors , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[19]  Geoffrey E. Hinton,et al.  Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.