Microwave imaging

Microwave imaging is a science which has been evolved from older detecting/locating techniques (e.g., radar) in order to evaluate hidden or embedded objects in a structure (or media) using electromagnetic (EM) waves in microwave regime (i.e., ~300 MHz-300 GHz). Engineering and application oriented microwave imaging for non-destructive testing is called microwave testing, see below. Microwave imaging is a science which has been evolved from older detecting/locating techniques (e.g., radar) in order to evaluate hidden or embedded objects in a structure (or media) using electromagnetic (EM) waves in microwave regime (i.e., ~300 MHz-300 GHz). Engineering and application oriented microwave imaging for non-destructive testing is called microwave testing, see below. Microwave imaging techniques can be classified as either quantitative or qualitative. Quantitative imaging techniques (are also known as inverse scattering methods) give the electrical (i.e., electrical and magnetic property distribution) and geometrical parameters (i.e., shape, size and location) of an imaged object by solving a nonlinear inverse problem. The nonlinear inverse problem is converted into a linear inverse problem (i.e., Ax=b where A and b are known and x (or image) is unknown) by using Born or distorted Born approximations. Despite the fact that direct matrix inversion methods can be invoked to solve the inversion problem, this will be so costly when the size of the problem is so big (i.e., when A is a very dense and big matrix). To overcome this problem, direct inversion is replaced with iterative solvers. Techniques in this class are called forward iterative methods which are usually time consuming.On the other hand, qualitative microwave imaging methods calculate a qualitative profile (which is called as reflectivity function or qualitative image) to represent the hidden object. These techniques use approximations to simplify the imaging problem and then they use back-propagation (also called time reversal, phase compensation, or back-migration) to reconstruct the unknown image profile. Synthetic aperture radar (SAR), ground-penetrating radar (GPR), and frequency-wave number migration algorithm are some of the most popular qualitative microwave imaging methods. In general, a microwave imaging system is made up of hardware and software components. The hardware collects data from the sample under test. A transmitting antenna sends EM waves towards the sample under test (e.g., human body for medical imaging). If the sample is made of only homogeneous material and is of infinite size, theoretically no EM wave will be reflected. Introduction of any anomaly which has different properties (i.e., electrical/magnetic) in comparison with the surrounding homogeneous medium may reflect a portion of the EM wave. The bigger the difference between the properties of the anomaly and the surrounding medium is, the stronger the reflected wave will be. This reflection is collected by the same antenna in a monostatic system, or a different receiver antenna in bistatic configurations. To increase the cross-range resolution of the imaging system, several antennas should be distributed over an area (which is called the sampling area) with a spacing less than the operating wavelength. However, the mutual coupling between the antennas, which are placed close to each other, may degrade the accuracy of the collected signals. Moreover, the transmitter and receiver system will become very complex. To address these problems, one single scanning antenna is used instead of several antennas. In this configuration, the antenna scans over the entire sampling area, and the collected data is mapped together with their antenna position coordinates. In fact, a synthetic (virtual) aperture is produced by moving the antenna (similar to the synthetic aperture radar principle). Later, the collected data, which is sometimes referred to as raw data, is fed into the software for processing. Depending on the applied processing algorithm, microwave imaging techniques can be categorized as quantitative and qualitative.