An Analytically Based Approach for Evaluating the Impact of the Noise on the Microwave Imaging Detection

In a realistic scenario, it is inevitable to have noise on the images due to the noise from the system's hardware, which results in producing inaccurate images. This paper presents an investigation on the impact of adding noises into the simulation for an Ultra-Wideband (UWB) Microwave Imaging (MWI) procedure based on the Huygens principle (HP). A comparison between uniform and Gaussian noises at different amplitudes is provided, with the aim of investigating the detection process for applications such as bone fracture detection. This is done using analytical simulations. To construct the electric field at the perimeter of the external cylinder, simulations have been run mimicking UWB signals transmitted onto a simulated cylindrical bone-mimicking phantom containing an inclusion with different dielectric properties. This field was simulated using MATLAB and generated a value for the electric field at frequencies between 3 and 5 GHz. To investigate the impact of noise on the detection capability, two types of common noises have been applied to the signal at different amplitudes. The resulting images have visually been compared and the imaging performance has also been analysed using an image quantification metric, signal-to-clutter ratio (SCR). The impact of noise on the detection capability was quantified using this image quantification metric.