The invention discloses a person
respiration non-contact detection method based on single-reflection schlieren. The method comprises the steps that a schlieren video
stream of a mouth and
nose area of a subject is collected through a single-reflection schlieren imaging device; the schlieren optical
system is configured to enable a
light spot of the
light source, which is gathered back through the reflector, to fall on the edge of an
entrance pupil diaphragm of the
imaging lens, and the physical edge of the
entrance pupil diaphragm is utilized to partially shield the gathered
light spot so as to form a virtual knife edge effect; performing frame-by-frame
processing on the schlieren video
stream, and in a
region of interest (ROI) below the mouth and the
nose, calculating a characteristic statistic representing
airflow disturbance intensity, namely a
respiratory airflow intensity
signal; performing
baseline drift removal and
frequency domain filtering
processing on the
respiratory airflow intensity
signal to obtain a standardized respiratory waveform; on the basis of the
breathing waveform, multi-dimensional
breathing parameters including the real-time
breathing frequency, the relative breathing depth and the inspiration and expiration ratio are calculated through a self-adaptive waveform
feature extraction algorithm; and analyzing transient characteristics and periodic laws of the respiratory waveform, and identifying
apnea, cough or sneezing events.