A Crack Monitoring Method of Bridge and Tunnel Structure Based on Abnormal Recognition of Digital Image
An abnormal image, bridge and tunnel technology, applied in character and pattern recognition, image analysis, image enhancement, etc., can solve the misjudgment of the background area that stays in the disease detection stage without cracks and diseases, and it is difficult to ensure the safe operation of bridge and tunnel structures. Dimension and other issues, to achieve the effect of improving computing efficiency and improving discrimination ability
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
specific Embodiment approach 1
[0022] Specific embodiment 1: A method for monitoring bridge and tunnel structural cracks based on digital image abnormality recognition described in this embodiment, the method includes the following steps:
[0023] Step 1: For the monitored bridge and tunnel structure, obtain the image data under the reference state, and use the single-dimensional Gaussian kernel function to construct a digital double-order multi-scale gain vector;
[0024] Step 2: According to the digitized double-order multi-scale gain vector obtained in Step 1, the kernel principal component analysis method is used to obtain the characteristic index of abnormal recognition of the digitized image under the reference state;
[0025] Step 3: Aiming at the feature index of digitized image abnormality recognition in the reference state obtained by solving in step 2, using the interval estimation method to calculate the digitalized abnormality discrimination threshold;
[0026] Step 4: Introduce the image data ...
specific Embodiment approach 2
[0029] Specific embodiment 2: This embodiment is a further description of a bridge and tunnel structural crack monitoring method based on digital image anomaly recognition described in specific embodiment 1. In the embodiment, the digital two-stage multi-scale method described in step 1 The method of constructing the gain vector is:
[0030] Step 11: According to the image data in the reference state in step 1, the single-dimensional Gaussian kernel function is used to calculate the double-order multi-scale matrix of each image:
[0031]
[0032] In the formula, Z k is the double-order multi-scale matrix of the kth image; X k is the double-order x-scale matrix of the k-th image; Y k is the double-order y-scale matrix of the kth image; f( ) is a custom function; I k is the gray value matrix of the k-th image; G is the single-dimensional Gaussian kernel; k is the serial number of the image data in the reference state; the symbol T indicates matrix transposition.
[0033] ...
specific Embodiment approach 3
[0047] Specific embodiment three: This embodiment is a further description of a bridge-tunnel structural crack monitoring method based on digital image anomaly recognition described in specific embodiment two. In this embodiment, the digitized The feature index solution method for image abnormality recognition is:
[0048] Step 21: Map the digitized double-order multi-scale gain vector in the reference state constructed in step 1 to a high-dimensional feature space, and construct the eigenvalue solution equation of its covariance matrix:
[0049]
[0050] C F ν=λν (8)
[0051] In the formula, is the kth in the reference state 0 The digitized double-order multi-scale gain vector of an image; is the kth in the reference state 0 The digitized two-order multiscale gain vector of an image Nonlinear representation in high-dimensional feature space; φ( ) is a nonlinear mapping function, which is expressed implicitly, and g h is the number of remaining 70% image data in...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More