CN-121998961-A - Visual detection method and system for product hidden cracking

CN121998961ACN 121998961 ACN121998961 ACN 121998961ACN-121998961-A

Abstract

The application discloses a visual detection method and a visual detection system for product hidden cracks, wherein the method comprises the steps of applying dynamic load to a sample, obtaining an original mosaic image of the sample under the dynamic load, generating a polarized intensity image group through a demosaicing algorithm, calculating and generating a polarized phase delay amount data layer corresponding to each frame of image, respectively calculating corresponding in-phase components and quadrature components, calculating phase locking amplitude values corresponding to each pixel point, combining the phase locking amplitude values to generate a two-dimensional phase locking amplitude distribution map, and generating a crack distribution image based on the two-dimensional phase locking amplitude distribution map to identify the hidden cracks of the sample. According to the application, static defect information of the sample crack is modulated into dynamic image information synchronous with the load, and the static defect information is separated from background noise through a lock-in amplification technology, so that the technical problem that the conventional detection means cannot accurately identify the micron-sized interface crack in a dynamic stress state is solved, and nondestructive visual detection of the product hidden crack is completed.

Inventors

YIN FANGTAO
WANG FAN
LU JINBO
ZHOU LONGSHENG
SUN MINGXIA
Xin Mingxin
WANG GUANGYU
WU YANNI
LI YULONG

Assignees

澳立奇科技股份有限公司

Dates

Publication Date: 20260508
Application Date: 20260131

Claims (10)

1. The visual detection method for the product hidden cracking is characterized by comprising the following steps: Acquiring a multi-frame original mosaic image of a sample under dynamic load, wherein each pixel point of the original mosaic image contains intensity information of one polarization direction; processing each frame of original mosaic image by a demosaicing algorithm to generate a polarized intensity image group which corresponds to each frame of image and contains at least four different polarization directions; Calculating and generating a polarization phase delay amount data layer corresponding to each frame of image based on the polarization intensity image group, and forming a polarization phase delay amount data cube by the polarization phase delay amount data layers of all frames according to time sequence; Applying a digital phase-locked amplification algorithm to each pixel point in the polarization phase delay data cube, and respectively calculating corresponding in-phase components and quadrature components; based on the in-phase components and the quadrature components, calculating a phase-locked amplitude value corresponding to each pixel point; Combining and visualizing phase-locked amplitude values corresponding to all pixel points to generate a two-dimensional phase-locked amplitude distribution map; a crack distribution image is generated based on the two-dimensional phaselock magnitude profile to identify a recessive crack of the sample.
2. The method for visually inspecting a product-implicit-crack according to claim 1, wherein the method for computing and generating the polarization phase retardation data layer corresponding to each frame of image based on the set of polarization intensity images comprises: and calculating a corresponding polarization phase delay amount data layer of each pixel point in the multi-frame polarization intensity image group through a polarization analysis algorithm.
3. The visual inspection method of product hidden cracking according to claim 2, wherein calculating the corresponding polarization phase retardation data layer of each pixel point in the multi-frame polarization intensity image group by using a polarization analysis algorithm comprises: Respectively calculating Stokes vectors of each pixel point; And calculating a corresponding polarization phase delay amount data layer of each pixel point in the multi-frame polarization intensity image group through Stokes vectors, and generating a time-varying polarization phase delay amount data cube.
4. The visual inspection method of product implicit cracking of claim 1, wherein the method of calculating the corresponding in-phase and quadrature components, respectively, includes: The in-phase component and the quadrature component are obtained by multiplying and integrating each polarization phase delay amount data layer in the polarization phase delay amount data cube with two quadrature reference signals respectively in time sequence.
5. The visual inspection method of product hidden cracking according to claim 1, wherein the method of combining and visualizing the phase-locked amplitude values corresponding to all the pixels comprises: and arranging a plurality of phase-locked amplitude values into a matrix, and linearly scaling the numerical values in the matrix to the standard image gray scale range for visualization.
6. The visual inspection method of product-implicit cracking of claim 1, wherein the method of generating crack distribution images based on two-dimensional phase-locked amplitude profiles includes: and generating a crack distribution image for identifying the defects according to the amplitude of each pixel point in the two-dimensional phase-locked amplitude distribution diagram.
7. The visual inspection method of product recessive cracking of claim 1, wherein the method for obtaining multi-frame original mosaic images of the sample under dynamic load comprises: And carrying out multiple exposure on the sample by an image sensor integrated with the four-way polarization filter array, and synchronously acquiring each frame of original mosaic image by one exposure.
8. The visual inspection method of product hidden cracks according to claim 1, wherein the electromagnetic vibration exciter is used for applying dynamic load to the sample, the dynamic load is periodic alternating load, and the waveform is sine wave.
9. The visual inspection method of product implicit cracking of claim 1, characterized in that the reference frequency of the digital lock-in amplification algorithm is synchronized with the frequency of the dynamic load.
10. A product-latent-crack visual inspection system for implementing the product-latent-crack visual inspection method according to any one of claims 1 to 9, characterized in that the product-latent-crack visual inspection system comprises: The image extraction module is used for acquiring multi-frame original mosaic images of the sample under dynamic load; The image preprocessing module is used for processing each frame of original mosaic image through a demosaicing algorithm to generate a polarized intensity image group which corresponds to each frame of image and contains at least four different polarization directions; the data processing module is used for calculating and generating a polarization phase delay amount data layer corresponding to each frame of image based on the polarization intensity image group, and forming a polarization phase delay amount data cube by the polarization phase delay amount data layers of all frames according to time sequence; The phase-locked amplifying module is used for applying a digital phase-locked amplifying algorithm to each pixel point in the polarization phase delay amount data cube and respectively calculating corresponding in-phase components and quadrature components; the calculation module is used for calculating the phase-locked amplitude value corresponding to each pixel point based on the in-phase components and the quadrature components; the visualization module is used for combining and visualizing the phase-locked amplitude values corresponding to all the pixel points to generate a two-dimensional phase-locked amplitude distribution map; and the crack detection module is used for generating a crack distribution image based on the two-dimensional phase-locked amplitude distribution diagram so as to identify the hidden cracks of the sample.

Description

Visual detection method and system for product hidden cracking Technical Field The invention relates to the technical field of product quality inspection, in particular to a visual detection method and a visual detection system for implicit cracking of a product. Background Currently, in the field of manufacturing high-end equipment such as aerospace, wind power generation and the like, graphene reinforced resin matrix composite materials are widely applied due to excellent performances such as high specific strength, high specific modulus and the like. However, these components are subjected to dynamic loads such as wind load, cyclic stress and mechanical impact for a long period of time in the use process, and are extremely prone to generating layered hidden cracks inside the material, particularly at the interface of the graphene reinforcing phase and the resin matrix. The cracks are positioned in the material, no visible trace exists on the surface of the material, but the fatigue resistance of the material can be obviously reduced under dynamic load, and finally, the disastrous consequences such as wind power blade fracture, aviation structural member failure and the like can be possibly caused. At present, the nondestructive detection technology of the hidden cracks has obvious limitations. Firstly, the traditional ultrasonic detection or X-ray detection is usually required to be carried out under off-line and static conditions, and the dynamic stress state born by a component in actual working cannot be truly simulated, so that in order to capture critical hidden cracks which only appear and are closed after unloading when being stressed, dynamic load capable of simulating the actual working condition needs to be applied to a sample. Meanwhile, in order to detect defects inside the material, an optical detection method based on static vision, such as polarized light imaging, is used, but when actually applied to a dynamic loading process, significant measurement errors are introduced. The reason is that although polarized light imaging can identify internal defects through material stress birefringence effects, it is difficult to effectively separate local, nonlinear, microscopic acoustic signals caused by interfacial cracking from global, linear optical background noise under the influence of dynamic loading. Therefore, an innovative method for realizing high-precision and high-signal-to-noise-ratio nondestructive detection on the hidden cracks of the interface of the graphene composite material in a dynamic stress state simulating an actual working condition is needed in the field. Disclosure of Invention In order to overcome the defects in the prior art, the invention provides the following technical scheme for achieving the purposes: the first aspect of the invention discloses a visual detection method for product recessive cracking, comprising the steps of obtaining a multi-frame original mosaic image of a sample under dynamic load, wherein each pixel point of the original mosaic image contains intensity information of one polarization direction; processing each frame of original mosaic image by a demosaicing algorithm to generate a polarized intensity image group which corresponds to each frame of image and contains at least four different polarization directions; Calculating and generating a polarization phase delay amount data layer corresponding to each frame of image based on the polarization intensity image group, and forming a polarization phase delay amount data cube by the polarization phase delay amount data layers of all frames according to time sequence; Applying a digital phase-locked amplification algorithm to each pixel point in the polarization phase delay data cube, and respectively calculating corresponding in-phase components and quadrature components; based on the in-phase components and the quadrature components, phase locking amplitude values corresponding to all the pixel points are calculated, and the phase locking amplitude values corresponding to all the pixel points are combined and visualized to generate a two-dimensional phase locking amplitude distribution diagram; a crack distribution image is generated based on the two-dimensional phaselock magnitude profile to identify a recessive crack of the sample. Further, the method for generating the polarization phase retardation data layer corresponding to each frame of image based on the polarization intensity image group calculation comprises the following steps: and calculating a corresponding polarization phase delay amount data layer of each pixel point in the multi-frame polarization intensity image group through a polarization analysis algorithm. Further, calculating the corresponding polarization phase delay amount data layer of each pixel point in the multi-frame polarized intensity image group through the polarization analysis algorithm comprises the following steps: Respectively calculating Stokes vectors of ea