Buried mines detection and classification : 2D-SAS processing definition and experimental results

The paper deals with the problem of buried mine detection and classification. GESMA (Groupe d´Etudes Sous-Marine de l´Atlantique) has been working on a new sonar concept for several months. A specific tool has been designed, made of a 2 m × 2m frame mounted on four legs to support and track sonar to permit 2D scanning of the seafloor. First, a non linear source is used and gathers data at sea, over different buried objects up to 50 cm, highlighting the interest of this sonar concept. The 2D sonar scan permits a 3D data matrix to be collected from which the detection and classification operations may be done step by step using successive 1D data (one sonar shot), multiple 2D images then multiple 3D displays. Presented results demonstrate the ability to classify and almost identify the buried object. Unfortunately, the use of the non linear effect may be risky for greater burial depth because the technique does not have good power efficiency. To overcome this problem, the use of a direct low frequency is envisaged. The high resolution needed for buried object classification is then achieved with the application of SAS processing. Since 2D scanning of the seafloor is made, we show that 2D SAS processing is made possible. The paper describes the new algorithm and presents experiments and results obtained after processing. The first experiments and processes give encouraging results on buried targets up to 1 m. The technique seems ideally suited for detecting and classifying buried mines.