Seabed classification of the South Tasman Rise from SIMRAD EM12 backscatter data using artificial neural networks

We have developed an automated method for sea‐floor classification for the South Tasman Rise, based on a SIMRAD EM12‐backscatter (13 kHz) mosaic and 47 sea‐floor samples. The samples have been divided into 3 distinct groups representing: (i) thick blankets of foraminiferal ooze; (ii) mixed sediments comprising sand/silt/mud (turbidites/chalk); and (iii) outcrops of metamorphic basement and volcanic rocks. A total of 515 sub‐areas, each measuring 32 × 32 pixels (∼4 km²) and representing the different seabed types, were extracted from the image from areas 128×128 pixels large, centred on the sample locations. The texture of the sub‐images was analysed by calculating grey‐level run‐length features, spatial grey‐level dependence matrices, and grey‐level difference vectors in four directions. A total of 100 samples for each class and 18 feature statistics were chosen to train an artificial neural network to recognise the textural attributes and their variability for each class. The performance of the network was evaluated first by classifying the image sub‐areas used for training, followed by classification of the remaining sub‐areas. Classification accuracies for the training samples for ooze, sand/silt/mud and basement rocks are 98%, 98% and 91%, respectively, and 91%, 83% and 84% for the test samples. Subsequently, we classified a total of more than 20 000 unknown image sub‐areas 4 km² large on the northern South Tasman Rise. The result agrees well with a visual geological interpretation of the sidescan mosaic as well as with a facies map of the area based on 3.5 kHz data. The success of a combined textural image and neural network analysis to classify a sidescan mosaic, in the presence of noise and processing artifacts, suggests a wide range of potential applications including the recognition of sediment textures and other objects in high‐resolution, shallow‐water backscatter images, and pattern recognition in remotely sensed geophysical images for mineral exploration.

Key words:

• backscatter data
• neural networks
• pattern recognition
• seabed classification
• South Tasman Rise

Notes

Present address: Nederlanske Aardolie Maatschappij B.V., Schepersmaat 2, Postbus 28000, 9400 Assen, Netherlands.