A method and device for monitoring the abnormality around a dam based on an unmanned aerial vehicle

Abstract

The invention discloses a method and device for monitoring the abnormality around a dam based on an unmanned aerial vehicle, belonging to the technical field of image processing. The device comprisesUAV unattended front-end equipment, a remote control center, a communication network, a data processing module, an algorithm module and a processing center. At first, that unmanned aerial vehicle (UAV) fly with the high-definition camera device, and the images of the designated position are collect and sorted; Each kind of image data is divided into training set and test set, and the training setis pretreated, and then the multi-scale fusion of convolution neural network is carried out to obtain the classification results. For the final classification results of each test set, the feature points of abnormal position are marked, and the training set and test set are divided again. Training the classification model of the training set image of the marked feature points; The test set is inputted into the classification model, and each feature point in each image is located in the abnormal situation. The invention improves the classification accuracy and the positioning accuracy, and solves the problem that the manual patrol inspection method is time-consuming and labor-consuming.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to an unmanned aerial vehicle-based abnormality monitoring method and device around a dam. Background technique [0002] At present, dam construction has become an important part of the national development strategy; the safe operation of dams provides guarantee for the development of the country and provides convenience for people's lives. In the dam inspection work, the commonly used method is manual inspection, through traditional methods such as visual observation, tapping and listening, and footsteps, or supplemented by simple tools such as hammers, brazing, and steel tapes to inspect the engineering surface and abnormal phenomena In order to judge whether there is an abnormality in the dam; the daily inspection frequency is normally once a week. In special circumstances, such as heavy rain, flood, and felt earthquake, the inspection will be carried out 24 hours a day ...

AI Technical Summary

Problems solved by technology

[0005] The present invention combines the defects existing in the prior art, based on the image classification and positioning method of multi-scale fusion on deep learning, aiming at the poor classification accuracy due to the existence of multi-scale and multi-view information in the existing dam anomaly classification and positioning method Due to the high problem and the difficulty in positioning because the abnormal situation usually appears in a small position in the picture, a UAV-based abnormality monitoring method and device around the dam is proposed;

Images