Device for realizing sitting posture detection

Abstract

The invention provides a device for realizing sitting posture detection. The device comprises a sitting posture calculation unit; a standard sitting posture registration unit used for determining whether registration succeeds or not according to the sitting posture result meeting a preset threshold value; a deviation unit for comparing the real-time sitting posture with the standard sitting posture used for carrying out deviation calculation on the acquired sitting posture information of the current frame and the registered sitting posture information; a unit for judging whether the deviation exceeds a threshold value or not used for determining that the current sitting posture is abnormal and further judging the state; a state continuous period judgment unit used for starting to accumulate the number of abnormal frames of the sitting posture when the sitting posture is judged to be abnormal and starting timing at the same time, and when the abnormal counting of the sitting posture exceeds a preset third threshold value, determining that the abnormal sitting posture is true and giving an alarm signal; and a unit for judging whether the sitting posture is standard or not used for starting to process the sitting posture calculation unit, comparing the real-time sitting posture with the standard sitting posture deviation unit and judging that the deviation exceeds a threshold value after the alarm signal is sent out, and canceling the alarm signal if the sitting posture is judged to be normal; and continuing to monitor the sitting posture information.

Description

technical field [0001] The invention relates to the field of head detection, in particular to a device for realizing sitting posture detection. Background technique [0002] With the continuous development of science and technology, especially the development of computer vision technology, head detection technology is widely used in various fields such as video surveillance and security, and the image feature extraction method has good recognition performance. Among them, sitting posture detection is also an important content in the field of recognition. In the prior art, it includes 1. Measuring the distance between the person and the sensor based on ultrasonic waves, laser lines, etc., and judging whether the sitting posture is standard or not. 2. Based on the depth sensor, collect information about people's sitting posture, extract key data to establish a sitting posture model, and use the sitting posture model to detect and judge the actual sitting posture. 3. Based on...

AI Technical Summary

Problems solved by technology

Because this method uses the face detection module and obtains the inclination angle of the face to judge the sitting posture, when the face is too large or blocked by other objects (such as wearing a mask, etc.), the complete face cannot be guaranteed to appear in the image , the failure of face detection leads to the failure of angle and human eye state detection, which eventually leads to the failure of sitting posture detection

[0014] 3. The patent CN107169453A is based on the depth sensor for sitting posture detection. It needs to use the depth sensor to obtain the depth image. The cost of the depth sensor is higher than that of the ordinary image sensor.

Images