Life below Water 2020



Research Projects | 101 Publications | 2 Patents

CoralSense: An Underwater Wireless Sensor Network for Monitoring Coral Reef Environment

Corals are one of the underwater species which has been immensely affected by rising sea temperatures and water pollution. Globally, the risk of coral bleaching (both moderate and severe) has increased at a rate of 4% per year, with 8% of reefs being affected by bleaching per year in the
1980s and 31% being affected in 2016.
NUST has committed to preserve the natural habitats, underwater ecosystems through research and technology, and has set an example of effective collaborations to materialize such efforts. NUST researchers, in collaboration with Umm Al-Qura University, KSA, are currently working on the
development of an Underwater Wireless Sensor Network (UWSN) lab to be the first specialized underwater sensor network lab, to support marine research and industry in the kingdom. The objectives of the collaboration are to design, develop and deploy a prototype UWSN for visual monitoring of coral reefs in the Red Sea.

Automatic Detection and Classification of underwater Man Made Object (MMOs)

The detection and classification of underwater objects is an important problem in a variety of naval and marine applications. Classification and detection of MMOs are traditionally carried out by a skilled human operator due to its difficulty. This is especially true for underwater environments where the variety of ambient noise present makes extraction and classification of relevant audio data a challenging task. The general approach for detecting targets is a two-tier process: 

  • Detection of possible naturally occurring and man-made objects
  • Classification into naturally occurring or man-made objects with a low detection rate of false alarms.

Researchers at Pakistan Navy Engineering College (PNEC), a constituent college of NUST based in Karachi, have proposed a solution that develops a characteristic feature vector-based technique in conjunction with optimized machine learning and data mining techniques to carry out automatic classification of underwater images using collected data. The project will be deployed in coastal areas of Pakistan to detect and analyze the under-water conditions of the coasts.

Development of a nondestructive approach to detect marine life in the freshwater of Pakistan

Researchers at NUST collaborated with Pakistan Museum of Natural History (PMNH) to collect fish data samples from different locations in Pakistan. The aim of the study was to develop a computer vision based algorithm for detection of fish species from the captured video frames by using background detection and foreground extraction and ultimately, classify fish species from collected imagery with the help of machine learning algorithms. The project also developed an easy to use Graphical User Interface (GUI) tailored for the dedicated task of generation and management of fish species detection. The project was deployed by PMNH for data collection and classification of local under-water species of Pakistan. The collaboration was an example of cross-sectoral deployment of NUST expertise for the restoration of environment and preservation of under-water life.