SDFT WITH KERNEL WINDOWING FOR 1D AND 2D SIGNALS

Abstract

The present work aims to get kernel values for all type of windows which can be cosine windows or non-cosine windows and use these values for different application in 1d and 2d application by using Parseval’s Theorem which is based on conservation of energy. Kernel windowing is special matrix value which are derived from different window func tion to indirectly apply the window function in different signal for different application of Signal Processing in 1d and 2d form. Kernel obtained in 1d windows is usually a row vector whereas it is matrix in case of 2d windows function for both separable window function or circular symmetric window function. Measurement of Capacitance can be done using many techniques where measurand quan tity which is capacitance converted to voltage, time, or frequency etc. here we are using conversion of capacitance to frequency using SDFT with KERNEL windowing which allows real time measurement, wide range of change of capacitance with good Accuracy. Relaxation oscillator is the most common used circuit for converting the capacitance in to the equivalent frequency, output of Relaxation Oscillator is square wave, so we use SDFT structure to extract the fundamental frequency of Square wave and then measure its frequency using Frequency Locked Loop which is based on sliding discrete Fourier transform. Allowance of Kernel windows can increase main lobe width and decrease side lobs depending on the window function are used to increase range of measured capacitance. Similarly, 2d Kernel windowing can be used in image processing for different application like Edge Detection and different Filtering application.