Igor V. The spline methods and tools of the analysis and syntheses for digital signals

In work are offered methods and tool of the signal processing of complex form, processing which classical methods not efficient. For instance, change the cubic polynom per cubic spline in filter Savitzky-Golay allows to increase window twofold when fixed to error or on 40% decrease the deviation an estimation when fixed window. For spline models of the signal, size and form the window are fixed naturally and are defined B-spline. However use splines in radiotechnikal systems requires the realtime and hardware tools and robust to noise. This requires the new original decisions of the methods and tools and algorithms. We Offered to build splines by means of convolution of continual fragment function (generative function). Received base spline consist of four fragments and have two continual derived. The Designed methods of the transformation base spline in interpolation pulses . This allows to avoid the decisions of interpolation equations. The Requirements to generative function correspond to the finiteimpulse response of FIR-filters. So task of the syntheses base splines correspond to task of the syntheses FIR-filter. Base spline possible consider the impulse response of the FIR-filter. This has allowed to estimate error of the processing signal in frequency domain in term of the digital signal processing and take into account the frequency characteristic a signal. Using splines, which are built with provision for a priori information on nature of the signal or his frequency characteristic allows greatly to raise accuracy of the description such signal. Unlike wavelet methods, which are founded on interpolation scheme, offered use the method least square (LS). Method of the calculation of LS spline is Received on uniform knots of unlimited dimensionality. It Is Designed fast algorithms spline-interpolations, filtering and compressions with use LS - splines. Practically create Wiener-Kolmogorov filters in class spline-function: LSS (Least Squares Spline) filters. Schemes Received robust towards white noise of frequency-times spline decompositions (LSS). The analytical LSS decomposition is executed for step function. The filters design of adaptive filtering. Adaptation is realized to account all three factors: modification the amplitude, scale and forms base spline. The Results marketed in digital device for programmed logic and software. Are Shown using received result for processing model and real radar, video and biometric signals.