Indoor wall crack detection method based on deep learning and image processing

Abstract

The invention discloses an indoor wall crack detection method based on deep learning and image processing, and relates to the technical field of image processing, and the method comprises the following steps: 1, constructing a crack image segmentation model, and introducing a jump layer in the crack image segmentation model; 2, importing a crack image, and performing pixel segmentation of a crack region on the crack image by using the crack image segmentation model; 3, calculating pixel length and width values of the segmented crack image; and 4, calculating a true crack value according to the calculated pixel length and width values. The jump layer is introduced into the crack image segmentation model, so that the detail segmentation of the crack image can be more accurate, and the function of high detection precision is realized. Meanwhile, the pixel length and width values of the crack image can be calculated, and then the real size of the crack is converted, so that the automatic detection task of the indoor wall crack can be completed, and the labor cost of related detection units is reduced.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to an indoor wall crack detection method based on deep learning and image processing. Background technique [0002] Due to the needs of safety, house decoration, etc., it is often necessary to check the interior walls, mainly to check whether there are cracks in the interior walls. The most primitive method is manual inspection, but this is time-consuming and labor-intensive for high-rise buildings, which is not realistic, and the artificial feedback of crack information can only be obtained by taking pictures, which will inevitably be not clear enough. [0003] With the development of technology, image information is obtained through photography, which replaces a large amount of manpower work. Based on the image information obtained by photography, there are also methods in the prior art that combine deep learning and image processing methods for automatic quantitative det...

AI Technical Summary

Problems solved by technology

There are still several problems in this way. On the one hand, the measurement of the real size of the crack is relatively lacking, and often only the pixel size of the crack is calculated, such as the pixel length and width; on the other hand, some technologies that can measure the real size of the crack require artificial Affixing high-precision calibration objects still requires more manual operations

Images