01-08-2023 дата публикации
Номер: CN116524190A
Принадлежит:
According to the real-time semantic segmentation method based on the improved double-flow feature extraction network provided by the invention, the real-time semantic segmentation network is constructed to perform semantic segmentation on the to-be-processed image data, and the real-time semantic segmentation network is a double-flow feature extraction network; according to the method, the input end feature maps are respectively from three feature maps with different resolutions in semantic paths, so that repeated calculation can be avoided, the parameter quantity of the model can be reduced, the relevance of feature information among the paths can be deepened, and the network model has a better multi-scale feature information fusion effect. By designing a spatial feature information enhancement module and changing spatial path input feature information, optimization of a traditional double-flow feature extraction network is completed, it is ensured that sufficient spatial feature detail ...
Подробнее