Construction site helmet wearing automatic detection method and system

Abstract

The invention discloses a construction site helmet wearing automatic detection method and system, and the method comprises the steps: obtaining a construction site monitoring video in real time, detecting all faces in the monitoring video, and obtaining a boundary frame of each face; tracking a bounding box of subsequent frames of faces in the monitoring video based on a face tracking algorithm, executing a face key point detection algorithm, detecting obtained face key points, and correcting a rectangular box of the face tracking algorithm by using a face bounding box of a current face key point detection result; solving a face Euler angle based on the detection result of the face key points and a PnP algorithm; detecting and tracking a human face, taking a safety helmet area with the best Euler angle obtained by the human face as a candidate area of the safety helmet, and judging whether the safety helmet exists in the candidate area of the safety helmet or not by using a safety helmet recognition algorithm. According to the invention, near-face detection and a safety helmet detection and recognition system are combined, so that the reliability of safety helmet recognition is improved.

Description

technical field [0001] The invention relates to the field of intelligent detection of safety helmets, in particular to an automatic detection method and system for wearing safety helmets at a construction site. Background technique [0002] The majority of safety incidents on production construction sites are related to worker violations. Falling, striking objects, collapses and hoisting damage are the four major types of fatal accidents. Also, falling and being struck by objects was the most common type of fatal accident, accounting for more than 68% of the total. However, due to the wide geographical distribution of the construction site, numerous equipment, complex environment, large amount of operations, and limited safety supervision, it is difficult to realize real-time whole-process safety management of the production and construction site. [0003] Recent related research mainly focuses on monitoring whether workers wear helmets by identifying the color information...

AI Technical Summary

Problems solved by technology

[0004] The above methods are mainly based on traditional computer vision methods, and some methods are based on color information for helmet recognition. The detection accuracy of these methods is low. The accuracy of the method based on deep learning proposed in this paper can be greater than 99%.

Moreover, one of the very meaningful tasks in helmet identification must determine its carrier (ie person). If only detecting a helmet is of little significance in many practical situations, and some methods even detect the carrier person when detecting a helmet, However, the current method mainly based on pedestrian detection cannot implement human identification.

Images