Unmanned aerial vehicle image information extraction system for flood disasters

Abstract

The invention provides an unmanned aerial vehicle image information extraction system, and the system is characterized in that the system comprises an unmanned aerial vehicle flight platform which is used for regulating and controlling an unmanned aerial vehicle to collect flood images, and dividing the flood images into a training set and a test set, a sample library construction module used for acquiring flood disaster images through the Internet and a search engine, and integrating the flood disaster images and the training set to construct a training sample library, a deep learning model building module used for training the sample data of the training sample library through a Yolo algorithm under a Tensorflow framework to generate a deep learning model, and a flood disaster extraction module used for importing the flood images of the test set into a deep learning model to obtain disaster information.

Description

technical field [0001] The invention relates to the field of UAV image technology, in particular to a flood disaster UAV image information extraction system. Background technique [0002] At present, as one of the main natural disasters, flood disasters cause huge casualties and economic losses every year. In the assessment of flood disaster loss, the information extraction of disaster-affected body is the premise of disaster loss assessment and arrangement of disaster relief and disaster reduction programs. It is difficult for traditional flood monitoring methods to provide disaster loss assessment data in a timely manner. In recent years, UAV remote sensing applications have become increasingly widespread, especially in the fields of natural disasters and emergency rescue, which have advantages that traditional methods cannot match. Facing flood disasters, this paper conducts research on flood disaster information extraction methods based on UAV remote sensing data and d...

AI Technical Summary

Problems solved by technology

Traditional flood monitoring methods are difficult to provide timely disaster loss assessment data

Images