Tumble detection method based on human body posture state judgment

Abstract

The invention discloses a tumble detection method based on human body posture state judgment. The tumble detection method comprises the following steps: extracting a target region of an image after binarization processing, extracting Zernike moment features, the height, the width, the aspect ratio, the target region area, the target region perimeter, the outline area, the outline perimeter, a frame difference maximum row and column characteristic variable and a moving object height feature variable of an object from the target region, respectively sending the extracted variables to two offline trained classifiers for judging moving conditions of a monitored person, and finally judging the real moving condition of the monitored person. According to the tumble detection method, an image processing technology is completely used to accurately distinguish posture states of walk, crouch and tumble, and the requirement of realtime processing on a hardware platform with lower performance can be met.

Description

technical field [0001] The invention relates to the field of computer vision information processing, in particular to a fall detection method based on human body posture state judgment. Background technique [0002] In recent years, with the continuous development of society, the demand for monitoring systems has developed from simple "watching" to "checking" and even "reporting". The key lies in how to use image processing technology to process the acquired video data, so intelligent The main function of the monitoring system is to extract the moving target from the continuous video images through the camera, identify and track the moving behavior of the moving target, and understand and describe its behavior. [0003] Among them, human body fall detection is an important part of human body motion behavior recognition, especially for the elderly living alone or individuals who are active in the wild alone. If no one knows about a fall, it may bring unimaginable consequences...

AI Technical Summary

Problems solved by technology

[0006] Similar technologies also have sensor-based fall detection, but this technology has the following disadvantages in actual use, including: it is necessary to wear additional hardware sensors to obtain human body movement information; the stability of communication between the acceleration sensor and other units of the system cannot be guaranteed; Accelerometer unit circuit components require regular charging; required sensor setup adds to system cost

[0007] The use of video image analysis is currently mainly based on complex mathematical modeling methods, which have some shortcomings in actual use, including: complex mathematical operations require high-performance hardware platform support to meet the system's real-time processing requirements, which undoubtedly reduces In terms of the definition and distinction of abnormal behaviors, the existing system does not detect specific abnormal behaviors, and the pertinence is weak, and it often regards squatting and other non-dangerous activities as normal human body movements. The behavior is defined as abnormal, and the false positive rate is high; in terms of practical algorithms, the dimensions of the features selected by most current schemes are often limited, and there is not enough feature dimensions to ensure the accuracy of detection

Images