Vision Based Underwater Environment Analysis: A Novel Approach to Estimate Size of Coral Reefs
Abstract
The purpose of our work is to detect and estimate the size of underwater coral reef from image frames using vision based technique. Deep learning architecture is proposed to detect the coral reefs. To extract the frames and build data set, real time underwater video is used. The size of reef is estimated using distance-based algorithm with detected coral information. A nonlinear function is formulated and optimized to get accurate reef size as pixels taken by it in image changes with distance. This algorithm can also be used for real-time analysis of the underwater environment. It is evident from the analysis results and comparison with actual data that the proposed method is accurate in estimating the area occupied by underwater coral reefs.
Collections
- Scholarly literature [9074]