Comparison of acoustic imaging methods and their variations with sequential implementations

In this paper we present our methods for acoustic imaging. Our goal was to develop a fast and robust method for ship-hull inspection in dirty harbor waters by using an acoustic imaging system. The SONAR system used had 5 MHz sampling frequency, 128 channels, a 1 MHz carrier frequency and a 125 kHz bandwidth. For creating acoustic images we used several techniques for quadrature demodulation and several methods for the beam forming. We described the variations for creating SONAR images and we implemented them in MATLAB and C++. For the algorithm development part we used flexible MATLAB implementation, and for the fast implementation of developed methods we used C++. The goal of this work was to measure the execution times of acoustic imaging methods and to find the optimum method for the particular implementation tool, MATLAB and C++, respectively. We compared different methods with different algorithmic approaches. Different optimal methods are determined for different implementation tools. Simulated and real-data experiments are presented in this work.