Development of a Model for Detection and Estimation of Depth of Shallow Buried Non-Metallic Landmine at Microwave X-Band Frequency

Detection and estimation of depth of shallow buried landmines using microwave remote sensing is a complex and computationally intensive task. Despite a lot of research to correctly locate and identify the buried landmines, and to estimate its depth using microwave remote sensing data which is essential for demining with minimal risk, a lot of uncertainties still exist. Therefore in this paper, an extensive study using a ground based X-band scatterometer for detection of shallow buried landmine and estimation of its depth has been carried out. An experimental setup consisting of a ground based scatterometer operating in microwave X-band (10 GHz, 3 cm) has been used to generate backscatter data in a grid of 24×24 and a series of experiments under laboratory conditions conducted using dummy landmines (without explosives) buried to different depths up to 10 cm in dry smooth sand. It is difficult to detect the buried landmine by visual inspection of the raw data; therefore a novel approach by fusion of image processing techniques with electromagnetic (EM) analysis has been evolved for detection and estimation of depth of the landmine. The raw data generated through the experiments was processed through a series of image processing steps and a 226 Tiwari, Singh, and Arora region of interest segmented using Otsu and maximum entropy based thresholding methods for further processing. A detection figure test has been proposed at this stage to reduce false alarms. Genetic algorithm (GA) with an electromagnetic (EM) model fusion has been proposed to estimate the depth after segmenting a suspected region containing the mine. The main advantage of the proposed model is that it does not have any requirement of separate training and test data set to train the optimizer and validate the results. Analyses of the results indicate that it is possible to segment suspected region of interest containing the landmines in data obtained in microwave Xband using either of the two thresholding methods. The depth of buried landmines estimated using the proposed GA optimized EM model was also found to be in good agreement with the actual depth. The proposed analysis is expected to be extremely useful in future in detection and estimation of the depth of landmines using satellite data in microwave X-band.