Construction high-altitude operation instability state detection method fusing spatial-temporal characteristics

Abstract

The invention discloses a construction high-altitude operation instability state detection method fusing spatial-temporal characteristics, and the method comprises the following steps: extracting the time motion characteristics of a detection target, and calculating the motion information of image sequence inter-frame pixels; extracting space positioning characteristics of the detection target, and fusing the space positioning characteristics with time motion characteristics to obtain space coordinate information of skeleton joint points of the building worker; and realizing high-altitude operation instability state detection according to the skeleton joint space position of the building worker. The method is realized through a neural network model formed by cascading a time sequence feature aggregation network model, a human body space positioning network model and a security state detection network model. According to the method, the spatial-temporal characteristics are fused, the spatial relation of human body skeleton articulation points on continuous frames of images is enriched by using the sequential relation of image pixels, and the construction aloft work instability state detection is realized by taking the gradient of the connection line of the skeleton articulation points as a threshold value. The method is high in intelligent level, and has a good application prospect and practical value.

Description

technical field [0001] The invention relates to the technical field of computer vision, including deep learning technology, and in particular, to a method for detecting instability of construction aerial work by integrating space-time features. Background technique [0002] The construction industry occupies an important position in my country's national economy, but compared with other industries, it also has high risks and frequent accidents. Although a large number of safety control measures have been formulated at the construction site, the current situation of construction safety is still grim. Among the many types of safety accidents, falling from heights is the most important type of accidents, accounting for more than half of the accidents. According to statistics, the unstable construction state such as leaning-over operations at heights can easily lead to falling accidents from heights. If the unstable state of construction workers can be detected and warned in a...

AI Technical Summary

Problems solved by technology

[0003] The purpose of safety status detection is to detect high-altitude dangers in a timely manner, but the current construction site is still dominated by manual detection, which has low efficiency and low frequency, and it is difficult to achieve all-weather and full coverage detection, and the written detection report submitted by manual detection is difficult to be detailed accurately reflect the test results, which is not conducive to the discovery and elimination of potential safety hazards

Images