A study on deep ocean image processing for object detection using pattern recognition

Deep ocean imaging and image processing techniques play a vital role in ocean engineering and scientific research. However, the underwater images usually suffer from low contrast, blurring, color fading, noise, and other distortions, due to these factors the identification of objects in the underwater image becomes difficult. In this work, we propose a multi-tiered underwater image processing system by incorporating various image enhancements, restoration and edge detection algorithms. The implementation of these mechanisms enhances the image quality by removing distortions and further this enhanced image is used to obtain patterns with which presence of various underwater species and objects can be analyzed