CN111928798B - A method for evaluating the quality of photoelectric imaging in phase deflectometry - Google Patents

Abstract

The invention relates to a phase deflection photoelectric imaging quality evaluation method. A projection device in a phase deflection system is used to project a sinusoidal fringe image used in detection to a standard measured object, and then a camera shoots the standard measured object with projection fringes. The image of the standard object to be measured is then processed by an algorithm to obtain a score of imaging quality. The higher the score, the higher the imaging quality, and the better the final detection effect of the actual object to be measured. The invention avoids the rework of the construction of the optoelectronic system, reduces the dimension of detection problem location, and reduces the time-consuming debugging of the quantity and phase deflection system; in the stage of construction of the optoelectronic system, the optoelectronic system is adjusted through the imaging quality evaluation standard until the imaging quality is reached. If the standard is met, the imaging quality is guaranteed at this time, which not only makes the optoelectronic system reach the ideal state in the construction stage, but also eliminates the imaging quality factor in the positioning of problems in the later detection stage (unsatisfactory detection results).

Description

A method for evaluating the quality of photoelectric imaging in phase deflectometry

technical field

The invention relates to the technical field of visual imaging detection, in particular to a phase deflectophotoelectric imaging quality evaluation method.

Background technique

Phase measurement deflectometry is based on the principle of phase detection to achieve object information acquisition. The projection device projects N+8 (N is generally 18) structured light fringes on the object to be measured in turn, including N/2 horizontal and vertical Gray code fringes and 4 horizontal and vertical sinusoidal fringes. The structured light is projected onto the surface of the object to be measured. The gradient change at the defect of the measured object causes the phase of the structured light to change, resulting in the deformation of the stripes. The CCD camera records the deformed structured light, and then obtains the phase information from the 26 frames of light intensity maps, and the camera takes pictures and transmits the images to the computer for images. After algorithm processing, an image with high contrast between defects and background is finally obtained.

In practical applications, due to the interference of ambient light, or the unreasonable settings of the camera pose and parameters, the quality of the fringes on the measured object captured by the camera will be unsatisfactory, resulting in unsatisfactory final detection results.

The traditional imaging quality debugging is based on the final test result as the evaluation standard, but there are too many factors that determine the final test result, and the test result is not ideal.

In the prior art, the phase deflection system can only be adjusted with the final detection result as the only evaluation criterion after the optoelectronic system, algorithm, etc. are all designed. This results in the adjustment being non-directional, making the adjustment work redundant and time-consuming.

SUMMARY OF THE INVENTION

The technical problem to be solved by the present invention is: to provide a photoelectric imaging quality evaluation method for phase deflectometry, so as to solve the problem that there is no quantitative and effective imaging quality evaluation method for phase deflectometry in the imaging stage, resulting in the necessity of final detection The result is a question of evaluation criteria.

The technical solution adopted by the present invention to solve the technical problem is: a method for evaluating the quality of photoelectric imaging by phase deflection, comprising the following steps:

A. Use the projection device in the phase deflection system to project the sinusoidal fringes used for detection to the standard measured object;

B. The camera shoots a standard object image with projected fringes;

C. The image of the standard object to be measured is processed by the algorithm to obtain the score of the imaging quality;

D. Judging whether the score meets the standard; if the score meets the standard, follow-up detection work; if the score does not meet the standard, adjust the parameters of the imaging system, and repeat steps 1) to 4).

Further, the ratio of the surface parameters of the standard object to be measured and the surface parameters of the actual object to be measured according to the present invention is 1:1~1:1.2.

Still further, the surface parameters of the standard measured object described in the present invention include the surface flatness, length, width and reflectivity of the standard measured object.

Furthermore, in the step C of the present invention, the higher the score, the higher the imaging quality, and the better the final detection effect of the actual measured object.

The invention uses the stripe image of the standard object to be tested to evaluate the imaging effect through the designed evaluation algorithm. When the evaluation score is low, the pose and parameters of the camera and the pose of the projection device are adjusted until a higher evaluation is obtained. Score, at this time, the imaging quality has become more ideal, that is, the imaging quality has been guaranteed; if the detection result is still unsatisfactory, the problem is not in the imaging quality, and the reason can be found in the algorithm and other aspects.

The beneficial effects of the present invention are:

1. Solve the problem that the phase deflection system does not have a quantitative and effective evaluation standard in the construction stage of the optoelectronic system. After the construction is completed, the optoelectronic system and algorithm should be adjusted based on the final inspection result, which will lead to the rework of the construction of the optoelectronic system. , and the reason for the unsatisfactory detection results is photoelectric or algorithm, it is impossible to determine this problem; it avoids the rework of photoelectric system construction, reduces the dimension of detection problem positioning, and reduces the time-consuming debugging of the phase deflection system. .

2. In view of the limitation of the traditional single use of the final inspection result as the evaluation standard, in the stage of building the optoelectronic system, the optoelectronic system is adjusted through the imaging quality evaluation standard until the imaging quality meets the standard, then the imaging quality is guaranteed. This not only makes the optoelectronic system reach an ideal state in the construction stage, but also eliminates the factor of imaging quality in the positioning of problems (unsatisfactory detection results) in the later detection stage.

Description of drawings

Fig. 1 is the flow chart of the method of the present invention.

Detailed ways

The present invention will now be described in further detail with reference to the accompanying drawings and preferred embodiments. These drawings are all simplified schematic diagrams, and only illustrate the basic structure of the present invention in a schematic manner, so they only show the structures related to the present invention.

As shown in Figure 1, a phase deflectophotoelectric imaging quality evaluation method, the imaging structure includes an area array camera, a projection device and a standard measured object.

The method includes: using the projection device in the phase deflection system to project the sinusoidal fringe image used in the detection to the standard measured object, and then the camera shoots the standard measured object image with the projected fringes, wherein the standard measured object surface is horizontal, and the length is long. , width and reflectivity are all close to the actual measured object; then the standard measured object image is processed by an algorithm to obtain an imaging quality score. The higher the score, the higher the imaging quality, and the better the final detection effect of the actual measured object.

The algorithm includes the following steps:

(1) Select a vertical sinusoidal fringe image, take a row of data, and form a one-dimensional array I;

w为图像宽度像素数；(2) Find the first derivative of the array I

w is the number of pixels of the image width;

；(3) Find the second derivative of the array I

;

(4) Obtain the coordinates of the zero-crossing point in I _d2 , which are the indices of the crest and the trough, and form these indices into a one-dimensional array I _d ;

(5) Find the absolute value of the difference between adjacent elements of I _d , and find the mean m_abs of the absolute value of the difference;

(6) Remove the elements whose absolute value of the difference between adjacent elements in I _d is less than m_abs, so as to remove the false peaks and false valleys caused by noise interference;

(7) Take the elements in I _d2 corresponding to I _d to form a new array I _{max_min} , find the mean value of I _{max_min} , and denote it as A';

(8) Use all elements in I _{max_min} greater than A' to form a new array I _max , and calculate its mean m_max; elements smaller than A' form a new array I _min , and calculate its mean m_min; calculate (m_max-m_min)/2 , denoted as B';

(9) Denote the mean value m_abs of the absolute value of the difference between the adjacent elements of I _d in step (5) as tmp1, calculate 3.14/tmp1, and denote it as C';

，记作D’；(10) Find the mean of the first 5 elements of the array I, denoted as tmp2, and calculate

, denoted as D';

(11) Bring I into the formula y=A+B*sin(C*x+D), and bring in A', B', C', D' as the initial values of A, B, C, D, Use the least squares method to find A, B, C, D;

(12) Bring I into the formula y=A+B*sin(C*x+D) to find the new array Y;

(13) Calculate the square of the difference between the corresponding elements between Y and I, and then take the average, and the result is recorded as S _tmp ;

,结果记作s_1;(14) Calculation

, the result is recorded as s_1;

(15) Randomly select 5 lines of data on the currently used vertical sinusoidal fringe image according to the above steps (1) to (14), and record the average of the 5 results obtained as the vertical imaging score s_v;

(16) In the same way, the data is taken in row units on the horizontal sinusoidal stripes, and the processing steps are the same as those on the vertical sinusoidal stripes, and the horizontal imaging score s_h is obtained;

,得到最终成像质量得分s,范围为0到1之间，数值越高代表成像质量越好；(17) Calculation

, get the final image quality score s, the range is between 0 and 1, the higher the value, the better the image quality;

Adjust the photoelectric system according to the evaluation score, including adjusting the pose and parameters of the camera, as well as the pose and parameters of the projection device, until the score reaches a preset score threshold (usually 0.8).

What is described in the above specification are only specific embodiments of the present invention, and various examples do not limit the essence of the present invention. Those of ordinary skill in the art can make modifications to the specific embodiments described above after reading the specification or variations without departing from the spirit and scope of the invention.

Claims (4)

1. A phase-shift-surgery photoelectric imaging quality evaluation method is characterized by comprising the following steps: comprises the following steps of (a) carrying out,

A. projecting sine stripes used in detection to a standard measured object by using a projection device in a phase deflection system;

B. shooting a standard measured object image with projection stripes by a camera;

C. processing the standard measured object image through an algorithm to obtain a score of imaging quality;

the algorithm comprises the following steps:

(1) selecting a vertical sine stripe image, and taking a line of data to form a one-dimensional array I;

(2) evaluating the first derivative of array I

w is the number of image width pixels;

(3) second derivative of array I

；

(4) ObtainTo obtain I_d2The coordinates of the middle zero crossing point are the indexes of the wave crest and the wave trough, and the indexes are combined into a one-dimensional array I_d；

(5) Finding I_dThe absolute value of the difference between adjacent elements, and solving the mean value m _ abs of the absolute value of the difference;

(6) removal of I_dElements with the absolute value of the difference between adjacent elements being smaller than m _ abs, so as to remove the pseudo wave crest and the pseudo wave trough caused by noise interference;

(7) get I_dCorresponding to I_d2The elements in (1) form a new array I_{max_min}To find I_{max_min}The mean value of (A) is denoted as A';

(8) by means of I_{max_min}All elements greater than A' in the array form a new array I_maxAnd calculating the average value m _ max, wherein the elements smaller than A' form a new array I_minAnd calculating the average value m _ min; calculating (m _ max-m _ min)/2, and recording as B';

(9) the value I obtained in step (5)_dThe mean m _ abs of the absolute values of the differences between adjacent elements is denoted tmp1, calculated 3.14/tmp1, denoted C';

(10) the mean of the first 5 elements of array I is found, denoted as tmp2, and calculated

Denoted D';

(11) substituting the formula I into a formula y = A + B sin (C x + D), substituting A ', B', C ', D' as initial values of A, B, C and D, and solving A, B, C and D by using a least square method;

(12) substituting the formula I into the formula Y = a + B sin (C x + D) to find a new array Y;

(13) the square of the difference between the corresponding elements between Y and I is calculated, then the average is taken, and the obtained result is recorded as S_tmp；

(14) Computing

The result is recorded as s _1;

(15) randomly selecting 5 rows of data on the currently used vertical sine stripe image according to the steps (1) to (14), and taking the average value of the obtained 5 results as a vertical imaging score s _ v;

(16) similarly, data is fetched on the horizontal sine stripe in a row unit, the processing steps are consistent with those on the vertical sine stripe, and a horizontal imaging score s _ h is obtained;

(17) computing

Obtaining a final imaging quality score s, wherein the range is 0 to 1, and the higher the numerical value is, the better the imaging quality is;

D. judging whether the score reaches the standard; if the score reaches the standard, performing subsequent detection work; and if the score does not reach the standard, adjusting the parameters of the imaging system, and circulating the steps A) -C).

2. The phase-bias-refraction-technique photoelectric imaging quality evaluation method as claimed in claim 1, characterized in that: the ratio of the surface parameter of the standard measured object to the surface parameter of the actual measured object is 1: 1-1: 1.2.

3. the phase-bias-technique photoelectric imaging quality evaluation method according to claim 2, characterized in that: the surface parameters of the standard measured object comprise the surface flatness, the length, the width and the reflectivity of the standard measured object.

4. The phase-bias-refraction-technique photoelectric imaging quality evaluation method as claimed in claim 1, characterized in that: in the step C, the higher the score is, the higher the imaging quality is, and the better the detection effect of the actual detected object is finally.

Images