Infrared technology defect reconstruction and feature extraction method based on multiplicative fuzziness

Abstract

The invention discloses an infrared technology defect reconstruction and feature extraction method based on multiplicative fuzziness, and the method comprises the steps of building a multi-region defect reconstruction model through an infrared thermal image sequence of a test piece to be detected, and reconstructing the defects in different space regions through variational Bayesian inference. Inthe collected data, the signals representing the internal defects are weak, and the noise may be introduced by parameter hypothesis during a defect reconstruction process, so that the internal defectinformation needs to be highlighted, the noise interference needs to be removed, the color features and the contour features of the images are extracted through a color extraction filter and an edge contour extraction filter, then the optimization processing is carred out on the defect color image, and finally the optimized color image and the contour features are fused through the inverse transformation, thereby clearly describing the defects of different space-time regions.

Description

technical field [0001] The invention belongs to the technical field of defect detection, and more specifically relates to a multiplicative fuzzy-based infrared technology defect reconstruction and feature extraction method. Background technique [0002] Infrared thermal imaging technology has been widely used in the automotive industry, shipbuilding industry, petrochemical industry and aerospace fields. Due to its features such as no need to directly contact the test piece to be tested, fast and efficient detection, and portability, it can effectively realize the nondestructive detection of damage. [0003] Infrared technology can be roughly divided into two categories, namely active heating and passive heating. For active heating, energy or heat is required to be artificially given to the test piece to be tested. According to the differences in heating methods, active heating can be divided into: optical excitation, electromagnetic excitation and mechanical excitation. O...

AI Technical Summary

Problems solved by technology

[0006] Secondly, the traditional defect method is mainly aimed at the defect characteristics of different regions in the same space, so the damage information in different spaces will be ignored, which will make a wrong judgment on the defect type of the material. In order to judge the defect situation of the material more accurately, The invention proposes an effective detection method, which can not only judge the damage of the surface space of the material to be detected, remove noise interference, but more importantly, obtain the damage situation of the inner space more accurately

Images