华南理工大学;SOUTH CHINA UNIVERSITY OF TECHNOLOGY;佛山市百步梯医疗科技有限公司;FOSHAN BAIBUTI MEDICAL TECHNOLOGY CO., LTD.
发明人:
MO, Hongqiang,莫鸿强,TIAN, Xiang,田翔
申请号:
CNCN2019/110287
公开号:
WO2020/114071A1
申请日:
2019.10.10
申请国别(地区):
CN
年份:
2020
代理人:
摘要:
A dual-microphone adaptive filtering method used for collecting body sound signals, the method comprising: using at least one primary microphone and one secondary microphone to collect signals, the primary microphone being used to collect body sound signals carrying noise, and the secondary microphone being used to collect environmental noise; performing identical high-pass filtering processing on the signals collected by the primary microphone and the signals collected by the secondary microphone; using a normalized least mean square algorithm on the primary microphone signals and secondary microphone signals following high-pass filtering processing to calculate adaptive filter weights and calculate error signals so as to filter out the environmental noise from the primary microphone signals; performing first low-pass filtering processing on the error signals so as to restore the body sound signals, thereby obtaining body sound signals outputted by the adaptive filtering algorithm. The present method may achieve the rapid convergence of filter weights, and may also avoid signal distortion; moreover, environmental noise interference may be quickly and reliably suppressed.L'invention concerne un procédé de filtrage adaptatif à double microphone utilisé pour collecter des signaux sonores corporels, le procédé comprenant les étapes consistant à : utiliser au moins un microphone primaire et un microphone secondaire pour collecter des signaux, le microphone primaire étant utilisé pour collecter des signaux sonores corporels porteurs de bruit, et le microphone secondaire étant utilisé pour collecter un bruit environnemental ; effectuer un traitement de filtrage passe-haut identique sur les signaux collectés par le microphone primaire et les signaux collectés par le microphone secondaire ; utiliser un algorithme des moindres carrés moyens normalisés sur les signaux de microphone primaire et les signaux de microphone secondaire suivant un traitement de filtrage passe-haut p