Target marking method and target recognition method for building facade damage detection

Abstract

The invention discloses a target marking method and a target recognition method for building facade damage detection, and a target recognition model can be quickly and finely adjusted through the target marking method. In the target recognition method, the target recognition model finely adjusted by the target marking method is adopted, so that the target recognition model adopted in the building facade damage detection process is more matched with the target building, and the accuracy and adaptability of the recognition process are improved; the problems that the difference of building facades of different buildings is too large, and the generalization ability and accuracy of a target recognition model are difficult to consider at the same time are solved.

Description

technical field [0001] The invention relates to the field of machine learning, in particular to a target labeling method and a target recognition method for damage detection of building facades. Background technique [0002] Traditional building facade damage detection usually requires inspectors to take pictures at close range to record and identify damage, but the efficiency of artificially identifying building surface damage is low, and the recognition quality is unstable, often requiring inspectors to spend a lot of time on inspection. Therefore, it is urgent to develop a fast and effective intelligent damage detection method. [0003] In order to improve the efficiency of damage detection, some intelligent detection methods based on machine learning have been proposed in recent years. The crack detection algorithm based on Faster R-CNN is used in the damage detection of bridge engineering, but it is sensitive to background noise, the calculation is slow, and it cannot ...

AI Technical Summary

Problems solved by technology

The crack detection algorithm based on Faster R-CNN is used in the damage detection of bridge engineering, but it is sensitive to background noise, the calculation is slow, and it cannot display damage in real time, and it can only be used to identify cracks and cannot identify other types damage

The Yolo algorithm has been used for real-time target recognition, but data labeling is often done in the background by developers who are not professional inspectors. The accuracy and efficiency are not ideal, and the actual application value is small.

In addition, the limitation of the input image resolution by the convolutional neural network makes it unable to deal with small features in high-resolution images, which may easily cause target recognition omissions, and there is still room for improvement.

Images