We consider in this paper the improvement of side-attack mine detection by performing confidence level fusion with data collected from vehicle-mounted forward-looking IR and GPR (FL-GPR) sensors. The mine detection system is vehicle based, and has both IR and FL-GPR sensors mounted on the top of the vehicle. The IR images and FL-GPR data are captured as the vehicle moves forward. The detections from IR images are obtained from the Scale-Invariant Feature Transform (SIFT) and Morphological Shared-Weight Neural Networks (MSNN) depending on target characteristics, and those from FL-GPR are derived from the FL-GPR SAR images through object-tracking.
Since the IR and FL-GPR alarms do not occur at the same location, the fusion process begins with each IR alarm and looks at the nearby FL-GPR alarms with confidences weighted by values that are inversely proportional to their distances to the IR alarm. The FL-GPR alarm with the highest weighted confidence is selected and combined with the IR confidence through geometric mean. An experimental dataset collected from a government test site is used for performance evaluation. At the highest Pd and comparing with IR only, fusing IR and FL-GPR yields a reduction of FAR by 26%. When the Hough transform is applied to reject the IR alarms that have irregular shapes, the fusion results provides a reduction of FAR by 35% at the highest Pd.
Near Infra-Red (NIR) offers enhanced contrast of man-made objects against vegetation. Shape detection algorithms for identifying side-attack mines in sequences of NIR imagery are described. These algorithms use morphological representations of features of the object in a network that learns features and classification simultaneously. A training set was constructed using NIR images of side attack mines. Testing sets were constructed using pairs of sequences of NIR images. Each pair of sequences contains a sequence containing a side attack mine and another sequence of the same scene with no side attack mine. Testing results from these sequences are presented.
Mathematical morphology is a field of knowledge and techniques involving the application of nonlinear image processing operations to perform image enhancement, feature extraction, and segmentation as well as a variety of other tasks. Morphological operations have previously been combined with neural networks to produce detectors that learn features and classification rules simultaneously. The previous networks have been demonstrated to provide the capability for detecting occluded vehicles of specific types using LADAR, SAR, Infrared, and Visible imagery. In this paper, we describe the application of morphological shared weight neural networks to detecting off-route, or “side attack”, mines. A pair of image sequences, both of the same scene, with and without a mine are presented to the system. The network then performs detection and decision-making on a per sequence basis.