In the construction process, wearing safety helmet can effectively reduce the injury caused by falling objects and hard objects. In order to detect the wearing of safety helmet during the construction process, a lightweight safety helmet wearing detection algorithm YOLOv5-CS is proposed and applied to the embedded equipment at the construction site. The algorithm is based on the improvement of YOLOv5 target detection algorithm. The improved CAShuffle module is used to form the backbone feature extraction network, which reduces the size of the original algorithm and ensures the detection accuracy of the algorithm; DIoU as the loss function, which makes the algorithm more conducive to detecting small targets such as hard hats; modify dataset, add the dataset on the basis of the original dataset and mark the safety helmet and the anchor frame of the person separately, and judge whether the worker wears the safety helmet by analyzing positional relationship of the identified anchor frame, use the algorithm combining K-means and genetic algorithm to predict the anchor frame of the dataset, and use Mosaic algorithm to enhance the data. The results show that the parameter quantity of YOLOv5-CS algorithm is 2.4M, less than one third of YOLOv5; The average precision decreased by only 4.2%; 640x640 pictures can be recognized at 31 frames per second on Jetson Nano, a embedded platform.
Based on the current characteristics of street lighting, the sensor data is extracted from the existing smart lighting devices, and the lighting failure prediction model of pigeon optimal BP network is built. The algorithm introduces magnetic navigation and landmark navigation in pigeon-inspired optimization into BP network, which solves the drawbacks of slow convergence and easy falling into local optimum of traditional network. The simulation results show that the algorithm can be applied to the prediction and diagnosis of urban street light illumination faults.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.