Mesoscale vortex intelligent identification method and system based on K8s

The invention provides a K8s-based mesoscale vortex intelligent identification method and system, and relates to the field of vortex identification. Comprising the steps of inputting a to-be-recognized image into a mesoscale vortex recognition model based on deep learning, performing feature extraction on the to-be-recognized image, and performing weighting based on a CBAM mixed attention mechanism; inputting the output feature map into ASPP, setting the expansion rate of expansion convolution, and replacing pooling branches in the ASPP with depth separable convolution to obtain a feature map output by the encoder; and inputting a feature map output by the encoder into a decoder, and calculating weights of receptive fields of different sizes based on an SKNet attention mechanism to realize intelligent recognition of the mesoscale vortex. According to the method, the cyclonic vortex and the anti-cyclonic vortex in the sea surface height image can be accurately identified, the model can more accurately segment