A fall detection method

Abstract

The invention discloses a fall detection method. The invention firstly trains the judgment threshold based on the training samples: after the training samples are subjected to median filter processing, then the acceleration signal vector amplitude is extracted, and the feature value is extracted: including the peak value of the acceleration signal vector amplitude, The difference between the minimum value and the maximum value of the acceleration signal vector amplitude, the standard deviation of the acceleration signal vector amplitude, and the relative angle change value; then based on the K-means clustering method to train the binary judgment threshold of each feature value; and then fall in real time During detection, the original acceleration information of the object to be detected is subjected to the same preprocessing as the training samples, and then four corresponding eigenvalues ​​are extracted, and then each eigenvalue is judged step by step to obtain the fall detection result. The invention can be used for real-time monitoring of the fall condition of the elderly group, and its computer complexity is low, and the real-time detection of the carrier's fall condition can be realized in the existing human body portable equipment.

Description

technical field [0001] The invention belongs to the technical field of computers, and in particular relates to a fall detection based on an acceleration sensor. Background technique [0002] In fall detection processing, the more common detection methods include: fall detection based on vision, fall detection based on sound, and fall detection based on sensors. [0003] Among them, the vision-based fall detection method mainly realizes the detection of the fall situation through the visual information of the human body movement process. In the vision-based fall detection process, the movement visual information of the human body is captured by the camera, and the Each frame of image is analyzed and processed to realize the recognition of the human body parts in the image. Finally, the judgment of the fall condition is realized by analyzing the sharpness of the change of the human body parts in different frames of images. However, fall detection based on video can only reali...

AI Technical Summary

Problems solved by technology

However, the detection efficiency and detection accuracy of existing sensor-based fall detection methods need to be further improved.

Images