Safe wearing detection method based on human body key points

Abstract

The invention discloses a safe wearing detection method based on human body key points, and the method comprises the steps of solving a detection problem of an occlusion person through employing a data enhancement method, increasing the number of occlusion samples through employing a random erasing method in a model training stage, and finally achieving the detection of a target person; correctingthe personnel target bounding box by using a symmetric space transformation network in a regional attitude estimation algorithm (RMPE), improving the accuracy of attitude estimation, solving the problem of redundant attitude by using parameterized attitude non-maximum suppression (Pose-NMS), obtaining the accurate attitude of the target personnel, extracting the key point information of the humanbody, and finally obtaining 17 key point coordinates; proposing a human body two-dimensional projection model, obtaining a head area by utilizing a seven-point positioning method according to the obtained coordinate information of the 17 key points, and obtaining a trunk area by utilizing a four-point positioning method; providing a ResNet50-TL-based safe wearing identification method, extractingfeatures from an obtained to-be-detected area, performing classification, obtaining the safe wearing condition of a worker, and realizing safe wearing detection. The method provided by the inventioncan effectively solve the safety wearing detection problem of the offshore drilling platform under the complex scene.

Description

technical field [0001] The invention belongs to the field of computer graphics and image processing, and relates to a safe wearing detection method based on key points of a human body. Background technique [0002] With the progress of science and technology and social development, in recent years, computer graphics and image processing technology has been more and more applied to people's production and life, especially in various engineering fields. The requirements are relatively simple and no contact is required, but the processing accuracy is very high, and it also has strong adaptability, so it can be widely used. Especially in the field of video surveillance, computer vision technology is widely praised for its low cost and high effectiveness, and the ability to customize different image processing algorithms for the functions of different surveillance systems. [0003] At present, safety behavior detection on industrial production and construction sites usually uses...

AI Technical Summary

Problems solved by technology

The current hard hat recognition algorithm is more effective in a single specific scene, but the background of the offshore platform is complex, and the posture of the staff is changeable, so it is difficult to guarantee a high recognition accuracy. less technical research

The problem of safety wear detection in simple scenes has been basically solved, but most of the detection methods are only suitable for simple scenes and single postures, and most of them only detect hard hats. Due to the influence of various factors such as posture and posture, it is difficult to perform accurate detection, and the recognition effect drops sharply. Safe wear detection in complex scenes has become an urgent problem to be solved.

Images