Aircraft inner cabin crack segmentation method based on adaptive non-local feature fusion

Abstract

The invention discloses an aircraft inner cabin crack segmentation method based on adaptive non-local feature fusion; the method comprises the steps: carrying out the down-sampling of an original image for multiple times, and obtaining a first low-level feature, a second low-level feature, a third low-level feature and a fourth low-level feature; sequentially performing multi-scale context extraction on the fourth low-level features, capturing long-distance dependency and adaptive fusion, and obtaining adaptive fusion features with a global dependency relationship; performing up-sampling on the adaptive fusion feature to obtain a first advanced feature; performing fusion and up-sampling on the third low-level feature and the first high-level feature to obtain a second high-level feature; performing fusion and up-sampling on the second low-level feature and the second high-level feature to obtain a third high-level feature; and performing up-sampling on the third advanced features to obtain a final feature extraction result. According to the method, through an adaptive feature fusion process, the global receptive field of the features can be captured, and the fusion problem of different scale features is effectively solved.

Description

technical field [0001] The invention belongs to the technical field of non-destructive testing, and in particular relates to a method for segmenting cracks in an aircraft inner cabin based on adaptive non-local feature fusion. Background technique [0002] With the vigorous development of emerging industries such as information science and artificial intelligence, computer technology is increasingly used in all aspects of national economic production such as industry, agriculture, and modern service industries. As an important branch of computer science, machine vision technology is based on image perception devices such as cameras, and enables computers to have the ability to observe, recognize and recognize the world by empowering them. Machine vision has been widely used in dangerous working environments that are not suitable for manual work or in scenarios where artificial vision is difficult to meet measurement or detection requirements. [0003] In the field of non-de...

AI Technical Summary

Problems solved by technology

Among them, the method based on image classification can well avoid the influence of low resolution and poor quality of the original image, but this kind of method can only classify whether there is a crack or not, and cannot classify the shape, size and location of the crack. Judgment; Compared with the method based on image classification, the recognition accuracy and granularity of the crack detection method based on target detection are greatly improved, but the important information such as the size and texture of the crack is still not available; in addition, the prediction results at the edge of the crack are not ideal It is also a common problem in segmentation tasks at present. The main reasons are too much downsampling of the network, inaccurate original labels, and lack of edge preservation strategies. Therefore, the structural crack detection based on deep learning needs to be further deepened and improved.

Images