B-spline 근사화 기반의 심전도 신호 압축

Abstract

PURPOSE: A B-spline approximation based electrocardiogram signal compressing method is provided to use a base function of a B-spline curve as a cubic B-spline, thereby enhancing the effectiveness of the compression. CONSTITUTION: A knot vector of a constant interval is set up and a B-spline base function is produced (S100). A control point is formed in the determined knot vector and the recovery data is produced (S200). A restoration error, which is the arithmetical difference between the original data and the recovery data of an electrocardiogram signal, is calculated (S300). Each root mean square (RMS) of the restoration error of each knot vector and each difference value of the average value are calculated (S400). If at least one value among each of the difference value exceeds the predetermined reference value (S500), a new knot vector is added in the middle of the exceeded knot vector section (S550). A new knot vector including the added knot vector is set up and the original data of the electrocardiogram signal is compressed by approximating the B-spline (S600). [Reference numerals] (AA) Start; (BB) End; (S100) Produce a knot vector of a constant interval is set up and a B-spline base function; (S200) Form a control point in the determined knot vector and produce recovery data; (S300) Calculate a restoration error, which is the arithmetical difference between the original data and the recovery data of an electrocardiogram signal; (S400) Calculate each root mean square (RMS) of the restoration error of each knot vector and each difference value of the average value; (S500) Each RMS value - an average value > a predetermined reference value?; (S550) Add a new knot vector in the middle of the exceeded knot vector section; (S600) Compress the original data of the electrocardiogram signal by approximating the B-spline according to the knot vector