Water environment pollution condition detection and evaluation device based on unmanned aerial vehicle aerial photography data

The invention discloses a water environment pollution condition detection and evaluation device based on aerial photography data of an unmanned aerial vehicle. The device comprises a data cleaning module, a semantic segmentation module and an offline image analysis module. The data cleaning module is used for carrying out efficient redundancy removal operation on a large batch of unmanned aerial vehicle aerial image data input by a user, reducing the overall time overhead and outputting a file list subjected to redundancy removal; the semantic segmentation module is used for carrying out pixel-level segmentation on riverways, floaters, abnormal discharge ports, bank slope farmlands, bank slope buildings and bank slopes in the input actual aerial video frame data or the redundancy-removed image data, and outputting target segmentation results and confidence coefficients; and the offline image analysis module performs differentiation processing and labeling on a segmentation result output by the semantic segmen