An automated change detection approach for mine recognition using sidescan sonar data

Abstract

This paper presents a new automated approach for the mine detection and classification (MDC) problem based on change detection techniques using sidescan sonar images. Adopting change detection techniques benefits this approach to recognize mine targets without training data or prior assumption required in traditional detection methods. In this approach, post-classification comparison is designed to detect the changes and the statistical information of pixel distribution is employed for change decision analysis. Specifically, because of the special characteristics of shadows in sonar images, shape and coarseness features are taken into account and play an important role in this method. This approach was successfully applied to two sets of bitemporal sidescan sonar images and the results are presented in this paper. The results prove the applicability of this approach for mine detection.

A Review of Underwater Mine Detection and Classification in Sonar Imagery

“…The pixels belonging to the same homogeneous regions are assigned the same labels. Segmentation is a popular technique to separate the highlight and shadow regions correlated with mines in sonar imagery [27][28][29][30][31][32][33][34]. This is because the pixels representing mines have higher values than the average pixel intensity in the image, while the pixels representing shadows created by mines have lower values.…”

“…To distinguish a mine, the detection step is needed. Some detection solutions utilise matched or template filters [17,28,41]. In simple algorithms, the templates are used for detecting highlight and shadow combinations.…”

“…Some also convolve templates with distinct areas in the image and employ a threshold to establish if the analysed area constitutes an MLO [17]. In this attempt, the template covers all possible MLOs but is able to recognise the bottom clutters [28]. Another template matching method generates a subspace of shapes with six vertices representing the MLO.…”

Markov-based approaches for ternary change detection between two high resolution synthetic aperture sonar tracks

“…The transition matrix (p(w s |w s−1 )), denoted as C and whose the unnormalized version is given (17), has been designed following certain guidelines:…”

“…Incoherent change detection methods only use the amplitude images (amplitude of the backscattered energy) to find changes. While symbolic or object-based approaches rely on the extraction of features of interest (ATR detection [15], [16], echo-shadow pairs [17], [18]) in both images before matching them by considering their local arrangement, imagebased methods [7], [19] only consider pixels intensity to detect potential changes by means of the computation of a difference image.…”