Aerial image building detection and segmentation method and device based on Mask R-CNN

Abstract

The invention relates to the technical field of artificial intelligence detection. The invention relates to the field of aerial photography, in particular to an aerial photography building detection and segmentation method and device based on Mask R-CNN. The method comprises the steps that firstly, aerial images of urban buildings are acquired, the contours of building objects in the aerial imagesare marked, a training set and test set data are established, and a training data set is enhanced in a non-random covering data enhancement mode; constructing an aerial photo building detection and segmentation network; training the network by using the training data set, and performing test and performance evaluation on the trained segmentation model through test set data to obtain a final aerial photography graph building segmentation model; and applying the obtained model to the building aerial image needing to be processed by the user to obtain a final building aerial image segmentation image. According to the method, a deep learning method is used, the speed and the efficiency are improved, and the segmentation accuracy and the robustness of the model are improved by applying a transfer learning and non-random masking data enhancement method.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence detection, in particular to a method and device for detecting and segmenting buildings in aerial images based on Mask R-CNN. Background technique [0002] With the rapid development of the national economy, the speed of urbanization in China is also getting faster and faster, and the number of various buildings in towns is increasing sharply. At the same time, the maturity of drone aerial photography technology has also made urban aerial images and videos Widely used in urban planning tasks. In the task of urban planning, 3D reconstruction in the task of building a digital city is the main trend of development, but there are various types of urban buildings and are constantly updated, such as urban villages and temporary factories on the outskirts of the city. Regularity and other issues, constructing a 3D digital model of a building is a task with a lot of work. In the 3D modeli...

AI Technical Summary

Problems solved by technology

[0004] For the scene of large-scale building segmentation, aiming at the problems of low detection and segmentation accuracy and high cost of buildings in aerial images, the object of the present invention is to provide a method and method for detecting and segmenting buildings in aerial images based on Mask R-CNN. device to realize high-precision, high-efficiency automatic detection and segmentation of buildings in aerial photographs

Images