OBSHCHESTVO S OGRANICHENNOJ OTVETSTVENNOSTYU "BI-ON EMG";ОБЩЕСТВО С ОГРАНИЧЕННОЙ ОТВЕТСТВЕННОСТЬЮ "БИ-ОН ЭМГ"
发明人:
KARIMOV, Vladimir Ruslanovich,КАРИМОВ, Владимир Русланович,IVANYUK, Natalya Mikhajlovna,ИВАНЮК, Наталья Михайловна,PONIMASH, Zakhar Alekseevich,ПОНИМАШ, Захар Алексеевич,SHCHEPANSKIJ, Valentin Evgenev
申请号：
RURU2017/000993
公开号：
WO2019/132692A1
申请日:
2017.12.27
申请国别(地区):
RU
年份:
2019
代理人:
摘要：
The invention relates to the field of gaming devices and medical technology, and more particularly to the rehabilitation of persons after limb loss or stroke. An EMG signal of a user is obtained by means of a myoelectric reading device. The EMG signal is processed by non-overlapping segmentation. For each segment obtained, a set of EMG signal features is formed on the basis of the square of dispersion of the EMG signal for gesture classification. On the basis of the resulting set of features, a neural network hash function generates a secondary feature space for data which is supplied to a metric classifier and describes a gesture. The set of EMG signal features of each segment is transmitted via a data transmission channel to a control system of a working mechanism. The gesture type is identified on the basis of the set of EMG signal features by using an artificial neural network. A control signal is generated on the basis of the identified gesture type and information from a gyroscope and an accelerometer. The control signal generated is transmitted to the working mechanism. Feedback is obtained from the control system of the working mechanism by obtaining information from external sensors. A vibrating motor is started up on the myoelectric device. EMG data can be output from the reading device to a face of the EMG reading device or to a screen of the working mechanism. The result is an increase in the accuracy of limb positioning and of decision making with respect to an action to be performed.L'invention se rapporte au domaine des dispositifs de jeu et des techniques médicales, notamment au rétablissement de patients ayant perdu une extrémité ou dans un état post-AVC. On reçoit un signal EMG de l'iutilisateur via un dispositif de lecture myoélectrique. On effectue un traitement du signal EMG par une segmentation sans chevauchement. Pour chaque segment obtenu, on génère un ensemble d'indices de signal EMG sur la base du carré de la dispersion du signal EMG pour c
