Ultra-low power consumption visual calculation heart rate and heart rate variability method

Abstract

The invention discloses an ultra-low power consumption visual calculation heart rate and heart rate variability method. The invention relates to the technical field of physiological information extraction. The method comprises the following steps: taking a visible light image or a near-infrared image as an input source; an input source acquires an RAW original image through a CMOS; iSP image optimization processing, image scaling and clipping, PCN face retrieval, target tracking, region mask definition and color space projection are carried out, normalization processing is carried out after Gaussian filtering, Butterworth filtering and trend filtering processing, filtering is carried out through a median filter, and heart rate index drawing is obtained through spectral analysis pulse wavedescription and video output; on the basis of ARMv7a, no DSP or neural network acceleration unit is provided, and a low-cost edge calculation unit is formed; band-pass filtering parameters can meet the requirement for heart rate and heart rate variability collection of all ages without calibration; the method has good resistance to ambient light and motion interference, and is suitable for visiblelight and infrared monochromatic images.

Description

technical field [0001] The invention relates to the field of medical technology for extracting physiological information, and more specifically relates to a method for visually calculating heart rate and heart rate variability with ultra-low power consumption. Background technique [0002] Edge computing provides the computing power needed in the Internet of Things, close to the data source, that is, the data generation end and the application edge end. For example, today's self-driving cars and electric vehicles are widely used in the Internet of Things technology. In the future, cars will be connected to the Internet, and a variety of rich applications for driving and individuals will be provided on the car. In this process, it is impossible to send all the collected data and applications that need to make real-time judgments back to the cloud computing center for calculation; and then return the processing results to the device side, such a calculation process will cause...

AI Technical Summary

Problems solved by technology

[0014] The purpose of the present invention is: in order to solve the Euler video amplification method in the prior art, the near-infrared monochromatic video image cannot be processed, and the corresponding passband parameters of the bandpass filter need to be individually adapted to limit its applicability. Suitable for edge computing applications, the present invention provides a method for ultra-low power consumption visual calculation of heart rate and heart rate variability

Images