Hybrid Deep Learning Framework for Video-Based Rainfall Estimation. (IMAGE)
Caption
A novel framework integrates urban surveillance video data with a two-stage AI pipeline: an enhanced random forest classifier detects rain streaks and selects key image regions, while a hybrid deep neural network combining depthwise separable convolution and gated recurrent units accurately predicts rainfall intensity in real time.
Credit
Environmental Science and Ecotechnology
Usage Restrictions
Credit must be given to the creator.
License
CC BY