Optical deformation measuring method based on self-adaptive grid and electronic equipment
阅读说明：本技术 基于自适应网格的光学变形测量方法及电子设备 (Optical deformation measuring method based on self-adaptive grid and electronic equipment ) 是由 黄建永 段晓岑 林峰 于 2019-09-11 设计创作，主要内容包括：本发明涉及一种基于自适应网格的光学变形测量方法和电子设备。其中测量方法包括：获取包含感兴趣区域的散斑图像,在感兴趣区域划分均匀网格,确定对应的系数矩阵的条件数随网格单元尺寸变化的L型曲线；以L型曲线拐点对应的尺寸作为单元尺寸下限,并以同等尺寸作为初始上限,在图像的第二区域范围划分均匀网格,该第二区域包含该感兴趣区域；以第二区域划分的均匀网格计算并更新位移场；根据更新的位移场确定感兴趣区域的应变梯度场,及根据应变梯度场确定单元尺寸上下限,通过应变梯度和单元尺寸的映射关系确定单元尺寸的分布。本发明测量方法通过自适应选定网格尺寸,减少大应变梯度位置由于形函数不匹配造成的误差及小应变梯度位置的随机误差。(The invention relates to an optical deformation measuring method based on an adaptive grid and an electronic device. The measuring method comprises the following steps: acquiring a speckle image containing an interested region, dividing uniform grids in the interested region, and determining an L-shaped curve of the condition number of a corresponding coefficient matrix along with the change of the grid unit size; dividing a uniform grid in a second area range of the image by taking the size corresponding to the inflection point of the L-shaped curve as a lower limit of the unit size and the same size as an initial upper limit, wherein the second area comprises the region of interest; calculating and updating the displacement field by using a uniform grid divided by the second area; and determining a strain gradient field of the region of interest according to the updated displacement field, determining upper and lower limits of the cell size according to the strain gradient field, and determining the distribution of the cell size according to the mapping relation between the strain gradient and the cell size. The measurement method of the invention reduces the error of the large strain gradient position caused by the mismatching of the shape function and the random error of the small strain gradient position by adaptively selecting the grid size.)
1. An optical deformation measurement method based on self-adaptive grids is characterized by comprising the following steps:
acquiring a speckle image containing an interested region, dividing a uniform grid in the interested region, and determining an L-shaped curve of the condition number of a coefficient matrix corresponding to the current uniform grid along with the change of the grid unit size;
determining the size corresponding to the corner of the L-shaped curve as a lower limit of the unit size, and dividing a uniform grid in a second area range of the image by taking the same size as an initial upper limit, wherein the second area comprises an interested area;
calculating and updating the displacement field by using a uniform grid divided by the second area;
determining a strain gradient field of the region of interest according to the updated displacement field, and determining the upper limit and the lower limit of the unit size according to the strain gradient field;
and determining the distribution of the cell size through the mapping relation between the strain gradient and the cell size according to the upper and lower limits of the cell size and the strain gradient field.
2. The method of claim 1, further comprising:
the grid is divided according to the distribution of the cell sizes and the maximum twist angle alpha is usedt＝max(2α0-αi) The grid quality at this time is calculated as an evaluation index, where α0Is a critical value, αiIs the angle between two adjacent sides;
if α ist＞α0Stepwise reduction of the upper size limit d1Until the grid quality meets the set requirements.
3. The method of claim 1, further comprising:
the steps recited in claim 2 are repeated until the upper limit of the cell size varies by less than 1 pixel or the number of iterations is greater than a specified value t.
4. The method of claim 1, further comprising constructing an error function based on the difference between the gray values of the corresponding points before and after the deformation and minimizing the error function before acquiring the speckle image including the region of interest:
the above formula is written in matrix form:
5. the method of claim 1, wherein: the unit shape is a polygon with an order of 1 to ∞.
6. The method of claim 1, wherein the mapping of the strain gradient and the cell size is represented by a step function, a linear function, or a high order polynomial.
7. An electronic device, comprising: a processor and a computer readable storage medium, having stored thereon a computer program, wherein the program, when executed by the processor, implements an adaptive mesh-based optical deformation measurement method according to any of claims 1-6.
Technical Field
The invention relates to the field of optical measurement, in particular to an optical deformation measurement method based on a self-adaptive grid and electronic equipment.
Background
The deformation measurement plays a crucial role in the fields of engineering monitoring, scientific research and the like, wherein the optical deformation measurement method is widely applied because the optical deformation measurement method does not need to be in contact with a measured object and is flexible and simple to use. By dividing a series of units connected by nodes, gray information contained in the speckle image units before and after deformation is compared, and the displacement field distribution with sub-pixel precision can be obtained by solving. Although the prior art method can obtain a more accurate calculation result in general, when a complex deformation is calculated, if a unit with a larger size is adopted, the shape function cannot accurately describe the current deformation, so that the interpolation result has a significant deviation from the actual result. At this time, the method can be improved by an encryption grid, and the encryption grid method is mainly divided into an h-type encryption method and a p-type encryption method. The former is achieved by reducing the cell size step by step, and the latter improves the accuracy by increasing the cell order.
In fact, only the positions with large deformation gradient need to be encrypted, and the positions with small deformation gradient can introduce new errors due to unit encryption: reducing the cell size results in a reduction in the amount of information contained within the cell, causing an increase in random errors; increasing the unit order amplifies the effect of noise, and both methods contribute to increased computational cost. However, most of the methods in the prior art adopt a global encryption processing method, and the selection of parameters such as cell size and the like depends on the experience of a user to a great extent, thereby limiting the further popularization of the optical deformation measurement method.
Disclosure of Invention
In view of the shortcomings of the prior art methods, the present invention provides an adaptive grid-based optical deformation measurement method.
According to an aspect of the present invention, there is provided an adaptive mesh-based optical deformation measurement method, comprising the steps of:
acquiring a speckle image containing an interested region, dividing a uniform grid in the interested region, and determining an L-shaped curve of the condition number of a coefficient matrix corresponding to the current uniform grid along with the change of the grid unit size;
determining the size corresponding to the corner of the L-shaped curve as a lower limit of the unit size, and dividing a uniform grid in a second area range of the image by taking the same size as an initial upper limit, wherein the second area comprises the region of interest;
calculating and updating the displacement field by using a uniform grid divided by the second area;
determining a strain gradient field of the region of interest according to the updated displacement field, and determining the upper limit and the lower limit of the unit size according to the strain gradient field;
and determining the distribution of the unit size through the mapping relation between the strain gradient and the unit size according to the upper and lower limits of the unit size and the strain gradient field.
In a further embodiment, the adaptive mesh-based optical deformation measurement method further comprises:
the grid is divided according to the distribution of the cell sizes and the maximum twist angle alpha is usedt＝max(2α0-αi) The grid quality at this time is calculated as an evaluation index, where α0Is a critical value, αiIs the angle between two adjacent sides;
if α ist＞α0Stepwise reduction of the upper size limit d1Until the grid quality meets the set requirements.
In a further embodiment, the adaptive mesh-based optical deformation measurement method further comprises:
and repeating the steps until the variation of the upper limit of the unit size is less than 1 pixel or the iteration number is more than a specified value t.
In a further embodiment, the adaptive mesh-based optical deformation measurement method further comprises, prior to acquiring the speckle image containing the region of interest,
constructing an error function according to the difference of the gray values of the corresponding points before and after deformation and enabling the error function to obtain the minimum value:
the above formula is written in matrix form:
in a further embodiment, the unit shape is a polygon, with an order of 1 to ∞.
In further embodiments, the mapping of the strain gradient and the cell size is represented using a step function, a linear function, or a higher order polynomial.
According to another aspect of the present invention, there is provided an electronic apparatus including: a processor and a computer readable storage medium having stored thereon a computer program which, when executed by the processor, implements an adaptive mesh-based optical deformation measurement method as described above.
The method can adopt various unit shapes and orders, and the mapping relation of the strain gradient and the unit size can adopt a step function, a linear function or a high-order polynomial.
Compared with the prior art, the invention has the beneficial effects that: the method can effectively improve the calculation precision of the complex deformation field, has no subjective dependence on parameter determination, and is high in robustness and simple to use.
Drawings
The invention is further illustrated with reference to the following figures and examples.
FIG. 1 is a schematic flow chart of example 1 of the present invention.
Fig. 2A is a speckle image before deformation in example 1.
Fig. 2B is a speckle image after deformation in
FIG. 3A is a strain gradient field in the range of 100 (pixel) ≦ x ≦ 300 (pixel), 100 (pixel) ≦ y ≦ 300 (pixel) in example 1.
Fig. 3B is the adaptive mesh partitioning result in
FIG. 4 is the variation of the error with the absolute value of the strain gradient under three different grid conditions in example 1.
FIG. 5 is a schematic diagram of a computer-readable storage medium of an embodiment of the invention.
Detailed Description
In order that the objects, technical solutions and advantages of the present invention will become more apparent, the present invention will be further described in detail with reference to the accompanying drawings in conjunction with the following specific embodiments.
According to the basic concept of the present invention, there is provided an adaptive mesh-based optical deformation measurement method, including: acquiring a speckle image containing an interested region, dividing a uniform grid in the interested region, and determining an L-shaped curve of the condition number of a coefficient matrix corresponding to the current uniform grid along with the change of the grid unit size; determining the size corresponding to the corner of the L-shaped curve as a lower limit of the unit size, and dividing a uniform grid in a second area range of the image by taking the same size as an initial upper limit, wherein the second area comprises the region of interest; calculating and updating the displacement field by using a uniform grid divided by the second area; determining a strain gradient field of the region of interest according to the updated displacement field, and determining the upper limit and the lower limit of the unit size according to the strain gradient field; and determining the distribution of the cell size through the mapping relation between the strain gradient and the cell size according to the upper and lower limits of the cell size and the strain gradient field. By self-adaptively selecting the grid size, the error of a large strain gradient position caused by mismatching of shape functions and the random error of a small strain gradient position are reduced, and the calculation precision of a complex deformation field can be effectively improved.
In the present invention, the region of interest refers to a region in which it is desired to obtain displacement field information.
According to an embodiment of the present invention, there is provided an optical deformation measurement method based on an adaptive grid, including the steps of:
acquiring a speckle image containing an interested region, dividing a uniform grid in the interested region, and determining a change curve of a corresponding coefficient matrix condition number along with the cell size;
determining the size corresponding to the corner as a lower limit of the unit size, and dividing a uniform grid in a second area range of the image by taking the same size as an initial upper limit, wherein the second area comprises the region of interest;
calculating and updating the displacement field by using a uniform grid divided by the second area;
determining a strain gradient field of the region of interest according to the current displacement field, and determining the upper limit and the lower limit of the unit size according to the strain gradient field;
and determining the distribution of the unit size through the mapping relation between the strain gradient and the unit size according to the upper and lower limits of the unit size and the strain gradient field.
Some examples of the inventionThe above method further comprises: the grid is divided according to the distribution of the cell sizes and the maximum twist angle alpha is usedt＝max(2α0-αi) The grid quality at this time is calculated as an evaluation index, where α0Is a threshold value (the threshold value can be one half of the corresponding internal angle when the polygon is a regular polygon, but is not limited thereto, and can also be 1/3, 3/4, etc.), αiIs the angle between two adjacent sides;
if α ist＞α0Stepwise reduction of the upper size limit d1Until the grid quality meets the set requirements. And repeating the steps until the variation of the upper limit of the unit size is less than 1 pixel or the iteration number is more than a specified value t.
In some embodiments, where the cell shape is a polygon, the order may be 1 to infinity; and the mapping relation between the strain gradient and the cell size is represented by a step function, a linear function or a high-order polynomial.
Wherein the error function is: function of the difference between the grey values of the corresponding points before and after deformation. The function may be constructed according to different correlation criteria including, but not limited to, SSD correlation criterion (sum of absolute values of difference of gray values after normalization), SSD correlation criterion (sum of square of difference of gray values), SATD correlation criterion (sum of absolute values of difference of gray values hadamard transform), or MSD correlation criterion (sum of mean squares of difference of gray values), MAD correlation criterion (sum of mean absolute values of difference of gray values).
For better understanding of the present invention, the following specific example is