CN110910318A - Weak contrast schlieren small ball center calculation method for comprehensive diagnosis light path quick automatic collimation system - Google Patents
Weak contrast schlieren small ball center calculation method for comprehensive diagnosis light path quick automatic collimation system Download PDFInfo
- Publication number
- CN110910318A CN110910318A CN201911001106.2A CN201911001106A CN110910318A CN 110910318 A CN110910318 A CN 110910318A CN 201911001106 A CN201911001106 A CN 201911001106A CN 110910318 A CN110910318 A CN 110910318A
- Authority
- CN
- China
- Prior art keywords
- image
- schlieren
- small ball
- background area
- center
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 18
- 238000004364 calculation method Methods 0.000 title claims description 14
- 238000000034 method Methods 0.000 claims abstract description 55
- 206010040925 Skin striae Diseases 0.000 claims abstract description 27
- 230000009466 transformation Effects 0.000 claims abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 13
- 238000005260 corrosion Methods 0.000 claims abstract description 11
- 230000007797 corrosion Effects 0.000 claims abstract description 11
- 230000000877 morphologic effect Effects 0.000 claims abstract description 10
- 238000003708 edge detection Methods 0.000 claims abstract description 5
- 230000002708 enhancing effect Effects 0.000 claims abstract description 4
- 239000011324 bead Substances 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims 1
- 238000004458 analytical method Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 5
- 238000005259 measurement Methods 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 239000008188 pellet Substances 0.000 description 3
- 238000000926 separation method Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 230000004907 flux Effects 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M11/00—Testing of optical apparatus; Testing structures by optical methods not otherwise provided for
- G01M11/02—Testing optical properties
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The application provides a method for calculating the center of a schlieren ball with weak contrast of a comprehensive diagnosis light path quick automatic collimation system, which can meet the requirements of the comprehensive diagnosis collimation system on the precision of the schlieren ball and the collimation efficiency of a light path. The method comprises the following steps: firstly, carrying out linear gray level transformation on an original image, enhancing the contrast ratio of a striae small ball target and a background area, and carrying out binarization on the linear gray level change image by utilizing a maximum inter-class variance method; secondly, digital morphological operations such as corrosion and expansion are used, so that a continuous connected domain is formed in the background area, and the largest connected domain is searched to be the background area. And then carrying out unequal operation on the digital morphology processing image and the background area image to obtain a schlieren small ball target. Finally, based on edge detection, the center and radius of the schlieren sphere are fitted using least squares.
Description
Technical Field
The application relates to a method for detecting a target with weak contrast on a plane.
Background
In a large laser device, in order to comprehensively verify the load capacity of a key optical component, assembly or system under high flux, a set of laser parameter comprehensive diagnosis system needs to be established, and the system is a multifunctional and high-precision laser parameter diagnosis platform which is used for precisely diagnosing the characteristics of laser beams output by the device and providing comprehensive and precise laser parameters for researching a frequency conversion assembly and related scientific and technical problems. Among all parametric diagnostic tasks, one of the most important tasks is the measurement of high dynamic range far field focal spots. For the measurement of the far-field focal spot with a high dynamic range, a schlieren method is generally adopted for measurement, namely, a main lobe and a side lobe are separately measured, the side lobe is measured by using schlieren light blocking, the main lobe is amplified, and the focal spot with the high dynamic range is measured by splicing.
According to the principle of schlieren method measurement, accurate main lobe and side lobe light spots, particularly side lobe light spots, need to be obtained, a schlieren sphere needs to be used for accurately shielding the center of a side lobe image, and therefore the light path needs to be collimated. In the light path collimation process, the light paths of all monitoring positions need to be collected and the positions of light beams need to be interpreted so as to determine the size and the method of light beam convergence. However, since the object to be monitored differs for each monitoring position, the image characteristics acquired at each monitoring position differ. For example: at the initial stage of light path adjustment, light beam perforation is needed, so that the acquired image is a circular pinhole image; in the adjustment process of the second half-path light path, the schlieren small ball needs to shield the side lobe center, and the collected schlieren small ball image has the following three characteristics: 1) the schlieren spheres are black, 2) the spheres are surrounded by bright areas, and 3) the schlieren sphere image is surrounded by black areas. Because the striae small ball is required to accurately shield the center of the side-lobe image (the error is less than 2 pixels), the main-lobe distribution image with the side-lobe center completely shielded and the detail of the area around the side lobe clear can be collected. If the shielding effect is poor, a saturated area appears in the small schlieren area, and accurate splicing cannot be completed. Therefore, a new method is required to be provided, under the condition of weak contrast, the center of the striae small ball can be accurately calculated, so that the center of the side lobe image can be accurately shielded after the beam convergence process of acquiring the side lobe image and centering and moving the striae XY motor is completed, and the important guarantee is provided for the automatic reconstruction of the final far-field focal spot image.
Disclosure of Invention
In order to meet the requirements of a comprehensive diagnosis collimation system on the precision of schlieren spheres and the collimation efficiency of an optical path, the application provides a weak-contrast schlieren sphere center calculation method of the comprehensive diagnosis optical path rapid automatic collimation system.
The technical scheme of the application is as follows:
firstly, carrying out linear gray level transformation on an original image, enhancing the contrast ratio of a striae small ball target and a background area, and carrying out binarization on the linear gray level change image by utilizing a maximum inter-class variance method;
secondly, digital morphological operations such as corrosion and expansion are used, so that a continuous connected domain is formed in the background area, and the largest connected domain is searched to be the background area.
And then carrying out unequal operation on the digital morphology processing image and the background area image to obtain a schlieren small ball target.
Finally, based on edge detection, the center and radius of the schlieren sphere are fitted using least squares.
The application has the following beneficial effects:
in the method, firstly, the image contrast is enhanced by using linear gray scale transformation, and the image is binarized; secondly, a background area is searched by using a BLOB technology, and then the background area of the schlieren globule is cut, so that the influence of the background area is eliminated; and finally, calculating the center of the schlieren sphere by using a least square method. Experimental results show that the method can calculate the center of the schlieren globule with the difference between the gray level of the globule area and the gray level of the background area larger than 10, the error between a circle fitting method and a true value is smaller than 2 pixels, and the requirement of a comprehensive diagnosis collimation system on the precision of the schlieren globule is smaller than 3 pixels. Meanwhile, the whole algorithm for calculating the center of the schlieren globule consumes little time, so that the time is saved for the whole collimation process, and the efficiency of the comprehensive diagnosis system on light path collimation is improved.
Drawings
Fig. 1 is a data processing flow of the present application.
Fig. 2 is a gray scale conversion curve.
FIG. 3 is a schematic diagram of a process of linear gray scale stretching; wherein, (a) is an original image, (b) is a direct binarization result of the original image, and (c) is a linear gray scale stretching result.
FIG. 4 is a schematic diagram of a process for obtaining background areas of schlieren spheres; wherein, (a) is the direct binarization result of the image after linear stretching, (b) is the isolated point in the schlieren sphere area, (c) is the result after digital morphology operation processing, and (d) is the schlieren sphere background image.
FIG. 5 is a schematic diagram of a process for obtaining a schlieren sphere target; wherein, (a) is the target image of the schlieren ball, and (b) is the search of the schlieren ball.
FIG. 6 is a flow chart of BLOB analysis.
FIG. 7 is a schematic diagram of a least squares circle fit; wherein, (a) is the detection circle profile, and (b) is the fitting circle.
FIG. 8 is a comparison of schlieren bead background images; wherein (a) is the result of 1 corrosion and expansion, and (b) is the result of 1 corrosion and 10 expansion.
FIG. 9 is a comparison of binarized images; wherein, (a) is the result of the unstretched binarization, and (b) is the result of the stretched binarization.
Detailed Description
As shown in fig. 1, the scheme of the present application comprises the following steps:
1) performing linear gray stretching on the schlieren small ball image;
2) carrying out binarization by using a maximum inter-class variance method;
σ(t)2＝ω1(t)ω2(t)[μ1(t)-μ2(t)]2(3)
in the formula: omega1(t)——C1The number of pixels, ω, contained in2(t)——C2The number of pixels, μ1(t)——C1Average gray value of all pixels in2(t)——C2Average gray value of all pixels in the array. Let T take values between 0 and maximum gray level, so that sigma (T)2The maximum value of T is the optimal threshold for the maximum between-class variance.
3) Performing mathematical morphological expansion and corrosion on the image;
4) searching a maximum connected domain as a schlieren small ball background area;
5) carrying out NOR operation on the schlieren ball background area and the digital morphology processing image;
6) negation of the NOR operation result;
in the formula: f. ofball(x, y) -schlieren-bead image, fmorph(x, y) -eroding the dilated image, fbg(x, y) -schlieren sphere background image.
7) Searching the maximum connected domain as a small ball target;
8) detecting edges;
9) and fitting the circle center and the radius by a least square method.
Some of the principles involved in the main steps of the present application are described below.
Principle of linear gray scale transformation:
the gray scale transformation is divided into linear gray scale transformation and nonlinear gray scale transformation, and the principle of the linear gray scale transformation is as follows:
the linear transformation is to transform the dynamic range of the brightness value of the original image to a specified range or the whole dynamic range according to a linear relation. The purpose of this is to enhance the contrast of parts of the original image, compress parts that are not of interest, stretch parts of interest, and change the visual effect of the image. In order to highlight the interested target or gray scale region and suppress the uninteresting gray scale region, the gray scale of the image is usually processed by piecewise linear transformation. Assuming that the minimum and maximum values of the division threshold are gL and gH, respectively, the gray scale transformation curve is as shown in fig. 2.
Setting the gray value (0-255) after Y bit conversion, wherein X is the gray value (0-4096) of the original image, and the conversion formula is as follows:
in the formula:
mean of the original image, σxThe standard deviation of the original image, and k is usually between 3 and 5, k is selected to be 5 in this experiment, and then gH is 302.
In order to enhance the image by gray scale transformation, it is important to select a suitable transformation interval, and gL is selected to be 0 in this experiment, and gH is obtained by formula (2).
Principle of the variance method between the maximum classes:
for larger targets, the maximum inter-class variance is generally adopted for image binarization, and the basic idea of the maximum inter-class variance method is as follows: dividing pixels in an image into C by a threshold value T according to gray values1And C2Two kinds, C1From the gray value of [0, T]Pixel composition of (B) C2From gray value at [ T +1, L-1]The inter-class variance between the two classes is calculated as follows:
σ(t)2＝ω1(t)ω2(t)[μ1(t)-μ2(t)]2(3)
in the formula: omega1(t)——C1The number of pixels, ω, contained in2(t)——C2The number of pixels, μ1(t)——C1Average gray value of all pixels in2(t)——C2Average gray value of all pixels in the array. Let T take values between 0 and maximum gray level, so that sigma (T)2The maximum value of T is the optimal threshold for the maximum between-class variance.
Least square fitting circle principle:
the least square-based circle fitting method is a method for approximating the laser spot contour by a circle according to the least square principle (the sum of the squares of the residuals is minimum) under the premise of obtaining the edge contour of the circular laser spot. Assuming the equation of a circle is (x-a)2+(y-b)2＝r2(4)
Here, the residual is taken as: epsiloni＝(xi-a)2+(yi-b)2-r2(5)
Wherein i belongs to E, and E represents the set of all boundaries; (x)i,yi) Is the boundary coordinates of the image.
The residual sum of squares function is:
and N is the number of the characteristic points participating in fitting calculation. According to the least squares principle, there should be:
order to
Wherein
Fitted circle center coordinate x0、y0And the radius r is expressed as:
it can be seen from the formula (9) that although the algorithm for detecting the center and radius of the striae globule derived from the circle fitting based on the least square principle is complex, each parameter can be calculated only by circulating the boundary point once, the time complexity is o (n), and the circle radius is calculated only once after the center parameters a and b are calculated, so the calculation speed of the whole algorithm is fast.
In the process of quickly and automatically collimating the light path of the comprehensive diagnosis system, accurately calculating the center of the schlieren sphere is the most important step for collimating the whole light path and is a key link of far-field focal spot measurement based on a schlieren method. Through the analysis of the schlieren small ball image, the collected schlieren small ball image has the following three characteristics: 1) the schlieren spheres are black, 2) the spheres are surrounded by bright areas, and 3) the schlieren sphere image is surrounded by black areas.
Aiming at the characteristics of the schlieren small ball image, the step of calculating the center of the schlieren small ball is as follows: 1) performing linear gray scale stretching on the image; 2) acquiring an image background using digital morphology; 3) acquiring a schlieren small ball target; 4) the striae globule center was calculated using a least squares circle fit.
The present application is further detailed by one embodiment in conjunction with fig. 1 below.
1. Linear gray scale stretching
After the light beam emitted by the relay light source is diverged by the schlieren illumination negative lens, the light beam is radiated to the side lobe CCD. The maximum gray scale of the CCD image can reach 4096 because the gray scale of the side-lobe CCD is 12 bits. However, after the light beam is transmitted in a long distance and is diverged by the schlieren illumination negative lens, the energy radiated to the side-lobe CCD is very low, the maximum gray value is only 360 and does not reach 1/10 of the dynamic range of the CCD, the gray value of the schlieren small ball area is about 230, the gray value of the schlieren small ball background bright area is about 245, the difference between the gray value of the schlieren small ball and the gray value of the surrounding bright area is less than 15, and for an image with very low contrast and small difference between the gray value of the target and the background, the small ball cannot be separated from the background area by using the maximum inter-class variance (OTSU).
In consideration of the difficulty of extracting the schlieren globule target after directly binarizing the original image, the embodiment uses linear gray scale transformation to transform the gray scale of the original image to 0-300, thereby realizing the enhancement of the image with weak contrast and obtaining good effect. The linear tone stretching results are shown in fig. 3 (c).
2. Acquisition of schlieren spherule background area using digital morphology
As shown in fig. 4(a), the image after the linear tone stretching is binarized by the maximum inter-class variance method as shown in fig. 3(c), and as a result, the image has the following characteristics: 1) the schlieren globule target is clearly seen, and the schlieren globule boundary is obvious; 2) the peripheral boundary of the background area is obvious, but the inside of the background area is in a flake shape, the background is not a finished connected domain, and honeycomb-shaped small holes with different sizes are distributed in the middle. In order to generate a complete connected domain image from the background region, the present embodiment uses digital morphology operation for processing. First, the binarized image of fig. 4(a) is subjected to 1 etching operation to remove discrete points isolated in the small schlieren area as in fig. 2 (b). And secondly, performing expansion operation on the image with the isolated points removed for 10 times, so that the background areas are connected into a whole to form a continuous connected domain. The results after 1 etch and 10 dilation digital morphological operations are shown in fig. 4 c). As can be seen from fig. 4(c), the background of the schlieren spheres is clear, and the schlieren spheres have no isolated discrete points and are surrounded by a completely connected white background area, which is very beneficial to realize the separation of the background area.
For the result image after the digital morphology operation processing as shown in fig. 4(c), the BLOB technique searches for the largest connected region in the image, which is also the largest white region, and the background image of the schlieren globule with the largest region is shown in fig. 4(d), so that the schlieren globule is surrounded in the largest connected region.
This embodiment uses BLOB analysis to search for the background area of the schlieren ball and the schlieren ball target, i.e. uses a chain code table and line segment table based method to search for multiple independent connected areas in the image. The principle of the BLOB analysis method is to extract characteristic parameters of an object through morphological processing and connectivity marker analysis, and to identify a target according to the parameters. FIG. 6 is a BLOB flow diagram.
The purpose of feature extraction is to extract the final schlieren sphere target. The premise of feature extraction is connectivity analysis, and a connected region in a picture must be found first in order to obtain the features of a target object.
Defining a structure:
the structure array ONEBLOB [ ] pOneBlob is defined for storing all connected regions. All connected domain searching steps in the image are as follows:
(1) finding the first pixel with the gray scale of 255 and the coordinate of (i, j), searching the XY coordinate (xpos, ypos) of the connected domain with the point as the starting point, the chain code table (code), and the chain code table length ChainCodeNum.
(2) The chain code table is converted into a segment table dot.
(3) The area of the connected component is calculated.
(4) And calculating the center coordinates xCenter and yCenter of the connected domain, and setting the connected domain represented by the line segment table to be 128 to represent that the connected domain is searched.
(5) The horizontal direction maximum size xLength and the vertical direction maximum size xLength of the connected domain are calculated.
(6) Turning to step (1), searching for the next connected component until all pixels with 255 gray levels are identified as 128.
(7) The connected domain with the largest area is calculated, usually the largest connected domain is the optical target, and in fig. 2(c) the largest connected domain is the schlieren globule background.
3. Obtaining schlieren small ball target
1) In order to separate the schlieren globule from the background, the morphological processing image (fig. 4(c)) and the globule background image (fig. 4(d)) are subjected to nor operation, namely, when the gray value in the schlieren globule background image is 255 and the area in the morphological image is 0 is selected as the schlieren globule target area; then, the color of the striae small ball target image is reversed, and the striae small ball target image is shown as fig. 5(a) and is expressed by the formula:
in the formula: f. ofball(x, y) -schlieren-bead image, fmorph(x, y) -eroding the dilated image, fbg(x, y) -schlieren sphere background image.
2) And (5) obtaining a final schlieren sphere target, wherein the connected domain which is searched for the largest area in the graph (a) in the graph 5 is the schlieren sphere, and the area of the schlieren sphere is calculated as shown in the graph (b) in the graph 5.
4. Circle fitting of striae globule centers using least squares
For fig. 7(b), Sobel operator is used for edge detection, the striae globule circular profile is obtained as shown in fig. 7(a), the circle center and the radius of the striae globule are fitted by using the least square method of formula (9), the circle center and the radius obtained by final fitting by using the circular profile are (614.65,681.60) and (132.75), and the result is shown in fig. 7 (b). The effects of the present application will be described below.
The calculation effect of the center of the schlieren ball is mainly analyzed from four aspects of image gray level stretching, schlieren ball background obtaining, circle center and radius calculation by a circle fitting method and schlieren ball center calculation repetition precision analysis:
4.1 image Gray level stretch analysis
Since the image contrast is low, if contrast stretching is not performed, the binarized image obtained by using the maximum between-class variance is, as shown in fig. 9(a), as a background bright region and a small ball cannot be distinguished, and a background region of a schlieren small ball is discontinuous, the background bright region and the small ball cannot be recognized. If the image after stretching is binarized by using the maximum inter-class variance method, the result is shown in fig. 9(b), and it can be seen from the figure that the background is a complete connected region, and the gray features of the small ball and the background bright region are clearly distinguished, so that the method is very beneficial to the acquisition of the background region and the separation of the schlieren small ball target.
4.2 schlieren globule background acquisition analysis
Only if the background area of the schlieren pellet is accurately obtained, the pellet target can be separated from the background. When the gray-scale stretched image is directly binarized, the background is not a continuous bright area, and a large number of discrete black pixels exist, and the method of erosion and expansion is generally adopted for the image. If one-time corrosion and expansion is carried out, although the continuity of the small schlieren ball area is good, a continuous background bright area cannot be generated, as shown in fig. 8(a), the method adopted in the embodiment is a method of carrying out 1-time corrosion and 10-time expansion, and as a result, as shown in fig. 8(b), it can be seen that a connected background area is generated after the 1-time corrosion and 10-time expansion treatment, the small balls are completely contained in the background area, and the small balls are completely standard black small balls, so that the separation and identification of the background and the small balls are very facilitated.
4.3 least squares fitting results analysis
The center and radius of the schlieren sphere were fitted using equation (9) for the circular profile obtained in fig. 7(a), the center and radius were obtained using the circular profile as (614.65,681.60) and 132.75, and the true center and radius were obtained using the calibration method as (614.5,681.5) and 132, where the error in the X direction was 0.15 pixels, the error in the Y direction was 0.10 pixels, and the radius error was 0.75 pixels. Therefore, the error between the circle fitting method and the true value is less than 2 pixels, and the requirement of the comprehensive diagnosis collimation system for the striae globule precision to be less than 3 pixels is met. Therefore, the algorithm can calculate the center of the striae small ball with the gray difference between the small ball and the background larger than 10, and after collimation is finished, the collimation precision of the target point reaches 10 micrometers and is smaller than the requirement of 30 micrometers of the target precision.
In the method of the embodiment, after the edge profile of the schlieren globule is detected, as shown in fig. 7(a), the circle center and the radius of the schlieren globule are fitted by using the least square method of the formula (9), because each parameter can be calculated only by circulating the boundary point once during circle fitting, the time complexity is O (n), and the time consumption of the whole algorithm is 0.8 second, the time is saved for the whole collimation process, and the requirement of the comprehensive diagnosis system on the rapid collimation of the light path is also met.
4.4 schlieren sphere center calculation repeat accuracy analysis
In order to verify the repetition precision of the calculation of the striae small ball center, the calculation results of the small ball center of 10 striae small ball images in the light path collimation process of different experiments are selected, and are shown in table 1.
TABLE 1 comparison of error between the calculated center of the schlieren sphere and the true center (unit: pixel) for this example
| No. | The fitted centre | The true centre | The centre error | The fitted radius |
| 1 | (614.65,681.60) | (614.5,681.5) | X:0.15Y:0.10 | 132.75 |
| 2 | (495.16,513.72) | (495.89,512.09) | X:-0.73Y:1.63 | 117.5 |
| 3 | (614.09,586.53) | (613.67,585.61) | X:0.42Y:0.92 | 134.5 |
| 4 | (507.56,338.39) | (507.20,337.50) | X:0.36Y:0.89 | 134.5 |
| 5 | (496.59,568.70) | (495.57,568.09) | X:1.02Y:0.61 | 116 |
| 6 | (489.61,490.91) | (489.10,490.91) | X:0.51Y:0 | 116.5 |
| 7 | (433.34,550.92) | (433.45,550.35) | X:-0.11Y:0.57 | 136.5 |
| 8 | (446.46,541.02) | (445.63,541.18) | X:0.83Y:-0.16 | 134 |
| 9 | (421.72,544.85) | (420.35,543.46) | X:1.37Y:1.39 | 134 |
| 10 | (449.08,547.76) | (448.97,547.22) | X:0.11Y:0.54 | 134 |
Table 1 lists the error comparisons between the calculated striae and calibration pellet centers using this example for 10 collimation experiments. As can be seen from the table, the center error XY direction is within 2 pixels, wherein the average value of the error in the X direction is 0.54, and the average value of the error in the Y direction is 0.72. This shows that the calculation repeatability of the striae globule center meets the requirement of the comprehensive diagnosis system for the striae globule center error within 3 pixels.
In summary, the application provides a striae globule center calculation method based on combination of gray scale stretching and digital morphology aiming at the characteristics of weak contrast of the striae globule image, black globule target and the globule target surrounded by the background region in the rapid alignment process of the comprehensive diagnosis system. Firstly, carrying out linear gray level transformation on an original image, enhancing the contrast ratio of a striae small ball target and a background area, and carrying out binarization on the linear gray level change image by utilizing a maximum inter-class variance method; secondly, 1 corrosion and 10 expansion digital morphology operations are used, so that the background area forms a continuous connected domain, and the largest connected domain is searched to be the background area. Then, the digital morphological processing image (fig. 4(c)) and the background region image (fig. 4(d)) are subjected to an arithmetic operation of inequality using the formula (10), and a schlieren sphere target is obtained. Finally, based on edge detection, the center and radius of the schlieren sphere are fitted using least squares. The algorithm can calculate the center of the schlieren globule with the difference between the gray level of the globule area and the gray level of the background area larger than 10, the error between the circle fitting method and the true value is smaller than 2 pixels, and the requirement of the comprehensive diagnosis collimation system for the schlieren globule precision smaller than 3 pixels is met. Meanwhile, the time consumed for calculating the whole algorithm of the center of the schlieren small ball is 0.8 second, so that the time is saved for the whole collimation process, the efficiency of the comprehensive diagnosis system on light path collimation is improved, and the method is a novel method for effectively calculating the center of the schlieren small ball.
Claims (2)
1. The method for calculating the center of the striae globule with weak contrast of the comprehensive diagnosis light path quick automatic collimation system is characterized by comprising the following steps of:
firstly, carrying out linear gray level transformation on an original image, enhancing the contrast ratio of a striae small ball target and a background area, and carrying out binarization on the linear gray level change image by utilizing a maximum inter-class variance method;
secondly, performing mathematical morphological expansion and corrosion operation on the binary image to enable a background area to form a continuous connected area, and searching the largest connected area as the background area;
then, carrying out NOR operation on the digital morphology processing image and the background area image to obtain a schlieren small ball target;
and finally, fitting the center and the radius of the striae globule by using a least square method on the basis of edge detection on the determined striae globule target.
2. The striae sphere center calculation method according to claim 1, wherein the specific processing steps are as follows:
1) performing linear gray stretching on the schlieren small ball image;
in the formula: gL and gH are respectively the minimum value and the maximum value of the segmentation threshold;
is the mean value of the original image; sigmaxIs the standard deviation of the original image; k is a constant and is taken as 3-5;
2) carrying out binarization by using a maximum inter-class variance method;
σ(t)2＝ω1(t)ω2(t)[μ1(t)-μ2(t)]2(3)
dividing pixels in the image after linear gray stretching into C according to gray values by using a threshold value T1And C2Two kinds, C1From the gray value of [0, T]Pixel composition of (B) C2From gray value at [ T +1, L-1]Pixel composition in between; then in the above formula: omega1(t) is C1The number of pixels, ω, contained in2(t) is C2The number of pixels, μ1(t) is C1Average gray value of all pixels in2(t) is C2Average gray value of all pixels in the image; let T take values between 0 and maximum gray level, so that σ (T)2The maximum T value is the optimal threshold value of the maximum inter-class variance;
3) performing mathematical morphological expansion and corrosion on the image;
4) searching a maximum connected domain as a schlieren small ball background area;
5) carrying out NOR operation on the schlieren ball background area and the digital morphology processing image;
6) negation of the NOR operation result;
in the formula: f. ofball(x, y) -schlieren-bead image, fmorph(x, y) -eroding the dilated image, fbg(x, y) -schlieren sphere background image;
7) searching the maximum connected domain as a small ball target;
8) detecting edges;
9) and fitting the circle center and the radius by a least square method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911001106.2A CN110910318A (en) | 2019-10-21 | 2019-10-21 | Weak contrast schlieren small ball center calculation method for comprehensive diagnosis light path quick automatic collimation system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911001106.2A CN110910318A (en) | 2019-10-21 | 2019-10-21 | Weak contrast schlieren small ball center calculation method for comprehensive diagnosis light path quick automatic collimation system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110910318A true CN110910318A (en) | 2020-03-24 |
Family
ID=69816114
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911001106.2A Pending CN110910318A (en) | 2019-10-21 | 2019-10-21 | Weak contrast schlieren small ball center calculation method for comprehensive diagnosis light path quick automatic collimation system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110910318A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111508017A (en) * | 2020-04-08 | 2020-08-07 | 中导光电设备股份有限公司 | Method and system for positioning mark center with weak contrast |
| CN111860616A (en) * | 2020-06-30 | 2020-10-30 | 中国科学院西安光学精密机械研究所 | General acquisition method for weak contrast collimation image target center of comprehensive diagnosis system |
| CN111882537A (en) * | 2020-07-28 | 2020-11-03 | 研祥智能科技股份有限公司 | Visual inspection method and system |
| CN113537303A (en) * | 2021-06-24 | 2021-10-22 | 中国科学院西安光学精密机械研究所 | Multi-optical target rapid classification and identification method based on template matching |
| CN114580522A (en) * | 2022-02-28 | 2022-06-03 | 中国科学院西安光学精密机械研究所 | Method for identifying multiple optical targets based on least square circle fitting method |
| CN115294486A (en) * | 2022-10-08 | 2022-11-04 | 彼图科技(青岛)有限公司 | Method for identifying violation building data based on unmanned aerial vehicle and artificial intelligence |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135416A (en) * | 2010-12-30 | 2011-07-27 | 天津普达软件技术有限公司 | Online image detecting system and method for bottle covers |
-
2019
- 2019-10-21 CN CN201911001106.2A patent/CN110910318A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102135416A (en) * | 2010-12-30 | 2011-07-27 | 天津普达软件技术有限公司 | Online image detecting system and method for bottle covers |
Non-Patent Citations (5)
| Title |
|---|
| 冯学智等: "遥感图像的增强处理", 《遥感数字图像处理与应用》 * |
| 孙浩晏: "基于机器视觉的指针式仪表读数识别系统研究", 《中国优秀硕士学位论文全文数据库（电子期刊）信息科技辑》 * |
| 朱秀昌等: "最大类间方差法", 《数字图像处理与图像通信》 * |
| 杨磊: "多段线性拉伸增强算法及其FPGA实现", 《红外技术》 * |
| 王拯洲等: "综合诊断系统多维度重构小孔光斑中心测量方法", 《红外与激光工程》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111508017A (en) * | 2020-04-08 | 2020-08-07 | 中导光电设备股份有限公司 | Method and system for positioning mark center with weak contrast |
| CN111508017B (en) * | 2020-04-08 | 2023-11-03 | 中导光电设备股份有限公司 | Method and system for positioning mark center with weak contrast |
| CN111860616A (en) * | 2020-06-30 | 2020-10-30 | 中国科学院西安光学精密机械研究所 | General acquisition method for weak contrast collimation image target center of comprehensive diagnosis system |
| CN111860616B (en) * | 2020-06-30 | 2024-05-14 | 中国科学院西安光学精密机械研究所 | General acquisition method for weak contrast collimation image target center of comprehensive diagnosis system |
| CN111882537A (en) * | 2020-07-28 | 2020-11-03 | 研祥智能科技股份有限公司 | Visual inspection method and system |
| CN111882537B (en) * | 2020-07-28 | 2023-12-15 | 研祥智能科技股份有限公司 | Visual detection method and system |
| CN113537303A (en) * | 2021-06-24 | 2021-10-22 | 中国科学院西安光学精密机械研究所 | Multi-optical target rapid classification and identification method based on template matching |
| CN113537303B (en) * | 2021-06-24 | 2023-01-06 | 中国科学院西安光学精密机械研究所 | Multi-optical target rapid classification and identification method based on template matching |
| CN114580522A (en) * | 2022-02-28 | 2022-06-03 | 中国科学院西安光学精密机械研究所 | Method for identifying multiple optical targets based on least square circle fitting method |
| CN114580522B (en) * | 2022-02-28 | 2023-08-11 | 中国科学院西安光学精密机械研究所 | Method for identifying multiple optical targets based on least square circle fitting method |
| CN115294486A (en) * | 2022-10-08 | 2022-11-04 | 彼图科技(青岛)有限公司 | Method for identifying violation building data based on unmanned aerial vehicle and artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110910318A (en) | Weak contrast schlieren small ball center calculation method for comprehensive diagnosis light path quick automatic collimation system | |
| CN108492272B (en) | Cardiovascular vulnerable plaque identification method and system based on attention model and multitask neural network | |
| CN116309537B (en) | Defect detection method for oil stain on surface of tab die | |
| CN113935998B (en) | Rubber and plastic part mottling detection method based on machine vision | |
| CN106846352B (en) | Knife edge picture acquisition method and device for lens analysis force test | |
| CN108520514B (en) | Consistency detection method for electronic elements of printed circuit board based on computer vision | |
| CN110415296B (en) | Method for positioning rectangular electric device under shadow illumination | |
| CN112883986B (en) | Static infrared target lamp identification method under complex background | |
| CN109766818B (en) | Pupil center positioning method and system, computer device and readable storage medium | |
| CN112001917A (en) | Machine vision-based geometric tolerance detection method for circular perforated part | |
| CN108168541B (en) | Improved sub-pixel star point centroid positioning method | |
| CN111695373B (en) | Zebra stripes positioning method, system, medium and equipment | |
| KR20180090756A (en) | System and method for scoring color candidate poses against a color image in a vision system | |
| CN112464829B (en) | Pupil positioning method, pupil positioning equipment, storage medium and sight tracking system | |
| WO2020019648A1 (en) | Machine vision positioning method | |
| CN115526889B (en) | Nondestructive testing method of boiler pressure pipeline based on image processing | |
| CN110738644A (en) | automobile coating surface defect detection method and system based on deep learning | |
| Gibbons et al. | A Gaussian mixture model for automated corrosion detection in remanufacturing | |
| CN110930425B (en) | Damaged target detection method based on neighborhood vector inner product local contrast image enhancement | |
| CN114170165A (en) | Chip surface defect detection method and device | |
| CN115100104A (en) | Defect detection method, device and equipment for glass ink area and readable storage medium | |
| CN108764343B (en) | Method for positioning tracking target frame in tracking algorithm | |
| CN116740053B (en) | Management system of intelligent forging processing production line | |
| CN116758044A (en) | Terahertz image quality evaluation method based on ResNet network | |
| CN116519640A (en) | Method for measuring surface glossiness of silica gel key based on machine vision system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200324 |