Construction site safety detection method based on YOLO algorithm

Abstract

The invention discloses a construction site safety detection method based on a YOLO algorithm. The method comprises the steps of acquiring data shot by a construction site by a monitoring system; labeling the acquired data, normalizing labeling information, forming a test data set, generating a training set and a test set in the data set, forming a training model based on a YOLOv4 algorithm and by taking CIoU as a frame regression loss function, importing the data set into the training model for training and testing, constructing the training model based on the YOLOv4 algorithm. target information can be extracted and recognized faster and more effectively, the detection speed and precision are improved, some target detection and recognition tasks can be well achieved, CIoU serves as a frame regression loss function, faster convergence speed can be effectively achieved directly by minimizing the normalized distance between a prediction frame and a target frame, and target detection and recognition efficiency is improved. And the regression is more accurate and faster when the regression and the target frame are overlapped or even included.

Description

technical field [0001] The invention relates to the technical field of safety detection, in particular to a construction site safety detection method based on the YOLO algorithm. Background technique [0002] With the continuous growth of the social population and the progress of the economic level, the housing and traffic roads in our country are becoming increasingly tense. This is a problem that we need to solve urgently. Therefore, the number of construction sites has also increased, so the safety of construction sites will be particularly prominent. , among them, real-time supervision of unsafe behaviors of workers on the construction site helps to improve project safety management level and reduce safety accidents; But at the same time, it is under severe work pressure, so there is an urgent need for a construction site safety detection method based on the YOLO algorithm to solve the above problems. Contents of the invention [0003] The present invention provides a...

AI Technical Summary

Problems solved by technology

[0003] The present invention provides a construction site safety detection method based on the YOLO algorithm, which solves the time-consuming, labor-intensive and low-efficiency problem of manual supervision of unsafe behaviors based on safety management personnel in the prior art, and the management personnel are under severe work pressure at the same time. question

Images