Method for real-time detection of falling from bed behaviors based on depth image information

Abstract

The invention discloses a method for real-time detection of falling from bed behaviors based on depth image information. The method includes the following steps: (1) a depth sensor acquiring a depth image of the indoor scene; (2) updating a tracking region of the depth image; (3) through a continuously changing offset dimension, extracting eight neighborhood differential characteristics of each pixel in the tracking region; (4) acquiring a head region; (5) positioning the center of the head region; (6) acquiring an above the waist region; (7) optimizing head positioning; (8) determining human body;(9) extracting height characteristics; (10) detecting falling from bed: adopting a well-trained falling from bed classifier to conduct classification detection on the height characteristics to acquire a falling from bed detection result. According to the invention, the method uses a random forest classifier to acquire the head region, and optimizes head positioning, which guarantee positioning accuracy, and also uses a support vector machine to detect falling from bed, which guarantees higher accuracy and robustness.

Description

technical field [0001] The invention belongs to the technical field of digital image recognition, and more specifically relates to a method for real-time detection of bed falling behavior based on depth image information. Background technique [0002] In both hospital care and home care, bed falls can cause serious injury to patients, the elderly, and infants. If the falling behavior can be detected in time, the damage caused by it can be greatly reduced. Therefore, the automatic detection of falling bed just seems particularly important. At present, there are three main detection methods for falling from a bed: the detection of falling from a bed based on wearable devices, the detection of falling from a bed based on pressure sensors, and the detection of falling from a bed based on video images. [0003] The bed fall detection method based on wearable devices judges the human body’s fall behavior by detecting the acceleration. The amount of calculation is small, but it n...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the above defects or improvement needs of the prior art, the present invention provides a real-time detection method for bed falling behavior based on depth image information, the

Images