Research Description

The ocean possesses vast treasures of ancient submerged remains. The last two decades have witnessed numerable ventures of underwater ancient discoveries. The underwater images observed as such provide only little data. Researchers further undertake challenges in unraveling the hidden remains using approaches to present more meaning to these raw images. The object detection methodology is most widely used to trace the submerged objects in terms of image that were captured from modern maritime Side Scan Sonar (SSS) device. The images are usually of low color contrast due to varying lighting conditions. Moreover, this leads to loss of object texture and hence difficult to retrieve an object present on SSS image. This prompts an enhancement preprocessing prior to feature extraction. A comparative analysis on the different enhancement techniques such as Histogram Equalization, Contrast Stretching, EMD with Optimal have been considered in enhancing the side scan sonar images and corresponding enhancement processing. Exploratory outcomes demonstrate that EMDW technique can acquire exact edge information and improved Visual Appearance. The Empirical Mode Decomposition with Optimal Weight (EMDW) picture upgrade procedure has been utilized to enhance the nature of sonar pictures. However, there is a scope in improvising the contrast and detailed analysis on the various contrast enhancement techniques are in progress. Some of the feature extraction methods investigated includes Support Vector Machine (SVM) and YOLO. These are prone to time consumption. This research is directed towards evolving an effective preprocessing technique for feature extraction using the above mentioned algorithms with updations and advancements on the same.

Publications:

Journal:

1.  Somasekar M, Sakthivel Murugan S, "Reduction of Artifacts and Edge Preservation of UnderwaterImages Using Deep Convolution Neural Network", Fluctuation and Noise Letters, Vol. 21, No. 4 (2022) 2250025, 2 July 2022, World Scientific Publishing Company.

2. Somasekar M, Sakthivel Murugan S, “Fusion-based approach for quality enhancement of Underwater images”, Journal of Environmental Protection and Ecology, Vol.22, Issue 4, September 2021, pp.1676-1687.

Conferences:

1. Somasekar M, Sakthivel Murugan S, “Feature Extraction of Underwater Images by Combining Fuzzy C-Means Color Clustering and LBP Texture Analysis Algorithm with Empirical Mode Decomposition ”, 4th International Conference in Ocean Engineering on Emerging Opportunities and Challenges- Chennai, India, 18th – 21st February, Chennai, India, 2018.

2. Sakthivel Murugan S, Somasekar M, "Segmentation of underwater acoustic images by FCM with EMD", 176th Meeting of the Acoustical Society of America, 5-9 November, Victoria, Canada, 2018.

3. Somasekar M, Sakthivel Murugan S, Pradeesha SK,  Padmapriya N, “A Survey on Fuzzy C-Means with Empirical Mode Decomposition for Underwater Image Enhancement”, International Conference on Ship & Offshore Technology - India 2017, IIT Kharagpur during December 7-8, 2017

4. Somasekar M,  Sakthivel Murugan S, Pradeesha SK, Padmapriya N, “Empirical Mode Decomposition by Weight Optimization for Amelioration of Underwater Images ”, OSICON-17 Conference held at NCESS, Trivandrum during 28th – 30th August, 2017.