CN106780425A - A kind of VVER steam generator heat-transfer pipe vortexes detecting system localization method - Google Patents

Abstract

The present invention relates to image procossing applied technical field, a kind of VVER steam generator heat-transfer pipe vortexes detecting system localization method is specifically disclosed.The method includes：1st, template demarcation is carried out to system；2nd, target image is gathered, and carries out interference removal treatment；3rd, under the conditions of low contrast, figure enhancing is carried out using histogram equalization algorithm；4th, the non-perpendicular imaging that enhanced image is had into angle to VVER steam generators photocentre and imaging surface carries out plane of delineation projection, and extracts image border；5th, the center of circle that new figure after projecting will be extracted in the template center of circle and step 4 is contrasted, and calculates side-play amount between the two；6th, the side-play amount that step 5 is obtained is fed back into equipment, it is run according to the side-play amount of feedback, after motion stops, repeat step 1 to step 4, until the side-play amount of calculating acquisition is in threshold range, end position error correction.The method, can overcome the positioning defect of existing alignment system, greatly improve the positioning precision of system.

Description

A kind of VVER steam generator heat-transfer pipe vortexes detecting system localization method

Technical field

The invention belongs to image procossing applied technical field, and in particular to a kind of VVER steam generators heat transfer Pipe eddy detection system localization method.

Background technology

Traditional equipment positioning method is positioned using feedback position of encoder ring, but equipment has tooth Wheel gap, the inherent variability such as the reciprocal return difference of screw mandrel cannot be eliminated, according to coaxial encoder feedback mode often without Method overcomes above-mentioned inherent variability, but using terminal encoder feedback mode exist again the stability of a system be difficult to it is up to standard, The characteristics of high to mechanical precision requirement.

The content of the invention

It is an object of the invention to provide a kind of VVER steam generator heat-transfer pipe vortexes detecting system positioning side Method, can complete to be accurately positioned VVER steam generator heat-transfer pipe vortex monitoring systems.

Technical scheme is as follows：A kind of VVER steam generator heat-transfer pipe vortexes detecting system positioning Method, the method specifically includes following steps：

Step 1, template demarcation is carried out to system；

VVER steam generator heat-transfer pipe vortex monitoring system Plays locating templates are chosen, the positioning is intercepted Template image, and calculate the center of circle for obtaining the template；

Step 2, collection target image, and to obtain image enter row metal high light it is reflective under the conditions of interference go Except treatment；

VVER steam generator heat-transfer pipe vortex monitoring devices are moved into target location, and collection is obtained Target image is obtained, the Image denoising algorithm in being processed using conventional images under the strong reflective disturbed condition of metal is right The target image of collection is processed；

Step 3, under the conditions of low contrast, carry out figure enhancing using histogram equalization algorithm；

Step 4, enhanced image there is into the non-perpendicular of angle to VVER steam generators photocentre and imaging surface Imaging carries out plane of delineation projection, and extracts image border；

Step 5, the center of circle that will extract new figure after projection in the template center of circle and step 4 are contrasted, and are calculated two Side-play amount between person；

Step 6, the side-play amount that step 5 is obtained is fed back into equipment, it is run according to the side-play amount of feedback, After motion stops, repeat step 1 to step 4, until calculating the side-play amount for obtaining in threshold range, terminates Site error is corrected.

Described step 4 specifically includes following steps：

Step 4.1, enhanced image there is into the non-perpendicular of angle to VVER steam generators photocentre and imaging surface Imaging carries out plane of delineation projection；

If known image is source figure, the camera coordinates corresponding to it are tied to the spin matrix of world coordinate system It is R_old, the Intrinsic Matrix of image is K_old；It is target figure to need to carry out the image that perspective transform obtains, it The spin matrix that corresponding camera coordinates are tied to world coordinate system is R_new, the Intrinsic Matrix of image is K_new.Then to visual field in an object point P, if its under world coordinate system coordinate be P_W, it is in source figure and target The homogeneous coordinates of corresponding pixel coordinate are respectively I in figure_old, I_new, then

Wherein, Z_c1And Z_c2Zs of the respectively object point P under the camera coordinate system corresponding to source figure and target figure Coordinate, Cpoint is three-dimensional coordinate of the video camera under world coordinate system.If the position of video camera is constant, Only attitude changes, then eliminate (P_w- Cpoint) can obtain

Wherein,

Wherein,It is the focal length under unit transverse direction mesh spacing,It is the focal length under the mesh spacing of unit longitudinal direction, u₀It is image principal point abscissa, v₀It is image principal point ordinate, K₀Can be regarded as being set to principal point coordinate (0,0) Intrinsic Matrix；

Then：

Both members be multiplied by simultaneously matrix K ' it is inverse, then：

Order：

I=K'^-1I_new

Then

New figure after must can projecting.

Described step 3 specifically includes following steps：

The probability that each gray level occurs in step 3.1, acquisition targeted graphical, and draw the histogram of artwork shape；

Step 3.2, calculating obtain the gray level cumulative distribution function of targeted graphical, and draw straight after equalization Fang Tu；

According to the probability that each gray level occurs, the gray level cumulative distribution function of targeted graphical is calculated, rounded And close gray level is normalized to, the gray value after each pixel is normalized is assigned to this pixel, draws Image after weighing apparatusization.

Remarkable result of the invention is：A kind of VVER steam generator heat-transfer pipe vortexes of the present invention Detecting system localization method, can overcome the positioning defect of existing alignment system, greatly improve the positioning of system Precision.

Brief description of the drawings

Fig. 1 is a kind of VVER steam generator heat-transfer pipe vortexes detecting system localization method of the present invention Flow chart.

Specific embodiment

Below in conjunction with the accompanying drawings and specific embodiment is described in further detail to the present invention.

As shown in figure 1, a kind of VVER steam generator heat-transfer pipe vortexes detecting system localization method, the party Method specifically includes following steps：

Step 1, template demarcation is carried out to system；

VVER steam generator heat-transfer pipe vortex monitoring system Plays locating templates are chosen, the positioning is intercepted Template image, and calculate the center of circle for obtaining the template；

Step 2, collection target image, and to obtain image enter row metal high light it is reflective under the conditions of interference go Except treatment；

VVER steam generator heat-transfer pipe vortex monitoring devices are moved into target location, and collection is obtained Target image is obtained, the Image denoising algorithm in being processed using conventional images under the strong reflective disturbed condition of metal is right The target image of collection is processed；

Step 3, under the conditions of low contrast, carry out figure enhancing using histogram equalization algorithm；

The probability that each gray level occurs in step 3.1, acquisition targeted graphical, and draw the histogram of artwork shape；

Step 3.2, calculating obtain the gray level cumulative distribution function of targeted graphical, and draw straight after equalization Fang Tu；

According to the probability that each gray level occurs, the gray level cumulative distribution function of targeted graphical is calculated, rounded And close gray level is normalized to, the gray value after each pixel is normalized is assigned to this pixel, draws Image after weighing apparatusization；

Step 4, enhanced image there is into the non-perpendicular of angle to VVER steam generators photocentre and imaging surface Imaging carries out plane of delineation projection, and extracts image border；

Step 4.1, enhanced image there is into the non-perpendicular of angle to VVER steam generators photocentre and imaging surface Imaging carries out plane of delineation projection；

If known image is source figure, the camera coordinates corresponding to it are tied to the spin matrix of world coordinate system It is R_old, the Intrinsic Matrix of image is K_old；It is target figure to need to carry out the image that perspective transform obtains, it The spin matrix that corresponding camera coordinates are tied to world coordinate system is R_new, the Intrinsic Matrix of image is K_new.Then to visual field in an object point P, if its under world coordinate system coordinate be P_W, it is in source figure and target The homogeneous coordinates of corresponding pixel coordinate are respectively I in figure_old, I_new, then

Wherein, Z_c1And Z_c2Zs of the respectively object point P under the camera coordinate system corresponding to source figure and target figure Coordinate, Cpoint is three-dimensional coordinate of the video camera under world coordinate system.If the position of video camera is constant, Only attitude changes, then eliminate (P_w- Cpoint) can obtain

Wherein,

Wherein,It is the focal length under unit transverse direction mesh spacing,It is the focal length under the mesh spacing of unit longitudinal direction, u₀It is image principal point abscissa, v₀It is image principal point ordinate, K₀Can be regarded as being set to principal point coordinate (0,0) Intrinsic Matrix；

Then：

Both members be multiplied by simultaneously matrix K ' it is inverse, then：

Order：

I=K'^-1I_new

Then

New figure after must can projecting；

Step 4.2, using self adaptation Canny threshold skirt extraction algorithms, extract the edge of new figure after projection；

Step 5, the center of circle that will extract new figure after projection in the template center of circle and step 4 are contrasted, and are calculated two Side-play amount between person；

Step 6, the side-play amount that step 5 is obtained is fed back into equipment, it is run according to the side-play amount of feedback, After motion stops, repeat step 1 to step 4, until calculating the side-play amount for obtaining in threshold range, terminates Site error is corrected.

Claims (3)

1. a kind of VVER steam generator heat-transfer pipe vortexes detecting system localization method, it is characterised in that：Should Method specifically includes following steps：

Step 1, template demarcation is carried out to system；

VVER steam generator heat-transfer pipe vortex monitoring system Plays locating templates are chosen, the positioning is intercepted Template image, and calculate the center of circle for obtaining the template；

Step 2, collection target image, and to obtain image enter row metal high light it is reflective under the conditions of interference go Except treatment；

VVER steam generator heat-transfer pipe vortex monitoring devices are moved into target location, and collection is obtained Target image is obtained, the Image denoising algorithm in being processed using conventional images under the strong reflective disturbed condition of metal is right The target image of collection is processed；

Step 3, under the conditions of low contrast, carry out figure enhancing using histogram equalization algorithm；

Step 4, enhanced image there is into the non-perpendicular of angle to VVER steam generators photocentre and imaging surface Imaging carries out plane of delineation projection, and extracts image border；

Step 5, the center of circle that will extract new figure after projection in the template center of circle and step 4 are contrasted, and are calculated two Side-play amount between person；

Step 6, the side-play amount that step 5 is obtained is fed back into equipment, its peace is run according to side-play amount of feedback, After motion stops, repeat step 1 to step 4, until calculating the side-play amount for obtaining in threshold range, terminates Site error is corrected.

2. a kind of VVER steam generator heat-transfer pipe vortexes detecting system according to claim 1 is positioned Method, it is characterised in that：Described step 4 specifically includes following steps：

Step 4.1, enhanced image there is into the non-perpendicular of angle to VVER steam generators photocentre and imaging surface Imaging carries out plane of delineation projection；

If known image is source figure, the camera coordinates corresponding to it are tied to the spin matrix of world coordinate system It is R_old, the Intrinsic Matrix of image is K_old；It is target figure to need to carry out the image that perspective transform obtains, it The spin matrix that corresponding camera coordinates are tied to world coordinate system is R_new, the Intrinsic Matrix of image is K_new.Then to visual field in an object point P, if its under world coordinate system coordinate be P_W, it is in source figure and target The homogeneous coordinates of corresponding pixel coordinate are respectively I in figure_old, I_new, then

$I_{old}=\frac{1}{Z_{c1}}{K}_{old}{R}_{old}(P_w-Cpoint)$

$I_{new}=\frac{1}{Z_{c2}}{K}_{new}{R}_{new}(P_w-Cpoint)$

Wherein, Z_c1And Z_c2Zs of the respectively object point P under the camera coordinate system corresponding to source figure and target figure Coordinate, Cpoint is three-dimensional coordinate of the video camera under world coordinate system.If the position of video camera is constant, Only attitude changes, then eliminate (P_w- Cpoint) can obtain

$I_{new}={\mathrm{zK}}_{new}{R}_{new}{R}_{old}^T{K}_{old}^{-1}{I}_{old}$

Wherein,

$z=\frac{Z_{c1}}{Z_{c2}}$

$K_{new}=\left[\begin{array}{ccc}\frac{f}{d_u}& 0& u_0\\ 0& \frac{f}{d_v}& v_0\\ 0& 0& 1\end{array}\right]=\left[\begin{array}{ccc}1& 0& u_0\\ 0& 1& v_0\\ 0& 0& 1\end{array}\right]\left[\begin{array}{ccc}\frac{f}{d_u}& 0& 0\\ 0& \frac{f}{d_v}& 0\\ 0& 0& 1\end{array}\right]=K^{\′}{K}_0$

Wherein,It is the focal length under unit transverse direction mesh spacing,It is the focal length under the mesh spacing of unit longitudinal direction, u₀It is image principal point abscissa, v₀It is image principal point ordinate, K₀Can be regarded as being set to principal point coordinate (0,0) Intrinsic Matrix；

Then：

$I_{new}={\mathrm{zK}}^{\′}{K}_0{R}_{new}{R}_{old}^T{K}_{old}^{-1}{I}_{old}$

Both members be multiplied by simultaneously matrix K ' it is inverse, then：

$K^{\′-1}{I}_{new}={\mathrm{zK}}_0{R}_{new}{R}_{old}^T{K}_{old}^{-1}{I}_{old}$

Order：

I=K'^-1I_new

Then

$I={\mathrm{zK}}_0{R}_{new}{R}_{old}^T{K}_{old}^{-1}{I}_{old}$

New figure after must can projecting.

3. a kind of VVER steam generator heat-transfer pipe vortexes detecting system according to claim 1 is positioned Method, it is characterised in that：Described step 3 specifically includes following steps：

The probability that each gray level occurs in step 3.1, acquisition targeted graphical, and draw the histogram of artwork shape；

Step 3.2, calculating obtain the gray level cumulative distribution function of targeted graphical, and draw straight after equalization Fang Tu；

According to the probability that each gray level occurs, the gray level cumulative distribution function of targeted graphical is calculated, rounded And close gray level is normalized to, the gray value after each pixel is normalized is assigned to this pixel, draws Image after weighing apparatusization.