Ray Vision
Ray Vision
Ray Vision
Ray Vision
Ray Vision
Ray Vision
Our ROV is equipped with a neural network-based target recognition system capable of detecting and identifying underwater organisms. The system uses bounding boxes to localize and classify the captured organisms, enhancing the ROV's capabilities for underwater exploration and research. The integration of biomimicry and artificial intelligence techniques offers a novel approach to underwater robotics, enabling efficient and accurate identification of marine life forms. This abstract summarizes the design, implementation, and performance evaluation of the ROV and its target recognition system in underwater environments.
With the continuous development of computer vision and the exploitation of marine resources, underwater biological detection has entered the public eye and has been applied in fields such as underwater robotics, underwater exploration, and marine research. However, autonomous high-precision detection of marine organisms is urgently needed. Compared to terrestrial detection scenarios, underwater imagery suffers from color shifts, low contrast, blurring, noise, and other quality issues, which may significantly reduce the usefulness of such images for target recognition. Therefore, when using such images for target recognition, the accuracy of target recognition will be greatly reduced.
Comming Soon