Underwater object identification and recognition with sonar images using soft computing techniques

SONAR is a device which is used to detect objects over the seabed using sound waves. Due to frequent changes in the oceanic weather conditions, water currents are produced. It causes changes in the underwater too. The change that has happened underwater can be determined by periodical monitoring. Change detection and object identification procedures are essential for understanding the underwater environment. This work focuses on the development of neural network based change detection along with Adaptive-Neuro Fuzzy Inference System (ANFIS) for object identification using underwater sonar images. The change detection algorithm is implemented using supervised classification technique such as Feed forward and pattern recognition network. The accuracy in detection of changes in the sonar image using pattern recognition network is 95.05% and that of feed forward network is 80.50%. Similarly, the accuracy of Adaptive Neuro-Fuzzy Inference System for object recognition is 85%. The proposed methodology is less complex and effective for sonar images compare to existing methodologies.