Research on Water Surface Environment Perception Method Based on Visual and Positional Information Fusion
摘要
The water surface environment characterised by complexity and variability, is heavily influenced by weather. To address this problem, this paper proposes a water surface environment perception network based on the fusion of visual and positional information, and proposes an encoder-decoder based semantic segmentation neural network for classifying the pixel points of the input image into three categories: water, sky and environment (obstacles).