abstract |
The present invention relates to a method for expanding a water image data set. The method includes the following steps: S1, acquiring an existing water image data set as a first data subset; S2, performing expansion processing on the image data in the existing water image data set Obtain a second data subset; S3, construct and train a generative adversarial network, and use the generative adversarial network to generate water simulation image data to obtain a third data subset; S4, combine the first data subset, the second data subset and the third data subset; The three data subsets are normalized and combined into an extended data set of aquatic image data. Compared with the prior art, the present invention can effectively increase the coverage of the water image data set, so that the data set has images of a larger number and wider scenes, and reduces the cost of constructing the data set. |