CN101963666A - Method for correcting intensity image distortion of streak tube imaging laser radar - Google Patents

Abstract

The invention relates to a method for correcting the intensity image distortion of a streak tube imaging laser radar, and belongs to the technical fields of photoelectronic imaging and image processing. In the method, a static image and a dynamic image of a target are subjected to filtering, wherein the luminance value of a streak tube fluorescent screen of the current streak data point on a static calibration line is shown as L0i, and a gray value of a pixel point of a point (x, yi) in an image to be processed is shown as Gxyi; the distance that the current streak data point is away from the static calibration line is farther, namely accelerating voltage subjected to the current streak data point is larger, the weakening of an intensity value of the current streak data point is worse, so the intensity image distortion of the current streak data point is corrected. The method has the advantages of small computation, simplicity and easy operation, is suitable for a laser radar system based on steak tube three-dimensional imaging and particularly a real-time data-processing system, and can reflect target properties really and provide accurate data for subsequent range information.

Description

The bearing calibration of a kind of striped pipe imaging laser radar intensity image distortion

Technical field

The present invention relates to the bearing calibration of a kind of striped pipe imaging laser radar intensity image distortion, belong to photoelectronic imaging and technical field of image processing.

Background technology

The laser three-dimensional imaging radar is international in recent years research focus, and it is the product that laser technology combines with Radar Technology, and in military affairs, there is vast potential for future development in fields such as space flight.Imaging laser radar based on the striped pipe is most promising a kind of in all laser three-dimensional imaging radar systems.In the imaging laser radar system based on the striped pipe, the range information that obtains target is the core of total system, also is the final purpose of whole imaging task.Stripe pattern is the most important Data Source that extracts target range information, and the quality that stripe pattern is handled is directly connected to the degree of accuracy that follow-up target range is extracted.In the actual imaging system, it has been generally acknowledged that the video screen fringe intensity image that CCD catches promptly is the power that has reflected the target return laser beam.Yet, fluoroscopic bright dark not only relevant with the density of accelerated electron beam, also relevant with the accelerating potential of electron beam, the different electron beams that arrive constantly have different accelerating potentials, the relative intensity information of so dynamic stripe pattern has just produced certain distortion, if with this view data as follow-up distance extraction, not only can not react target really, more directly influenced follow-up range information extraction work, made range information extract and be based upon on the inaccurate data basis.

Summary of the invention

The objective of the invention is in order to solve in the prior art owing to the extraction that intensity is not proofreaied and correct the range information that causes as data is based upon on the inaccurate data basis, the bearing calibration of a kind of striped pipe imaging laser radar intensity image distortion is proposed, this method is to fringe intensity image weighted, method is simple, be specially adapted to real-time striped pipe three-dimensional imaging laser radar system, guaranteed that range information extracts the accurate of data.

The objective of the invention is to be achieved through the following technical solutions.

The bearing calibration of a kind of striped pipe imaging laser radar intensity of the present invention image distortion, its concrete steps are:

1) to the still image M of target _sAdopt least square method to carry out static datum line and demarcate, obtain the ordinate y of the i bar static demarcating line of single pixel wide _0i, all ordinates of static demarcating line are y ₀₁, y ₀₂... y _0i... y _0m, wherein i is 1～m, m is the fringe number of stripe pattern;

2) to the dynamic image M of target _dCarry out the border and keep the class smothing filtering, the image that obtains is designated as M _D0

3) according to the ordinate y of the static demarcating line that obtains ₀₁, y ₀₂... y _0i... y _0mAt filtered dynamic image M _D0The corresponding surely straight line of last corresponding coordinate subscript, demarcating the pending image of finishing is M _d';

4) measure the difference of target range distance as requested, set the predeterminated voltage value V of striped pipe _p, the luminous required minimum voltage value of striped tube fluorescent screen is V in the characteristic parameter of striped pipe _d

5) a semi-transparent semi-reflecting glass column is added the magnitude of voltage that scanning voltage is imaged as the point of a series of even distances and notes this series of points, the magnitude of voltage between any 2 of these a series of dibblings and the ratio between the distance value value that obtains of averaging is designated as r;

If x is that current fringe number strong point is in pending image M _d' in horizontal ordinate, y _iFor current fringe number strong point in pending image M _d' in ordinate, the scanning voltage value of the striped pipe of then current fringe number strong point correspondence is formula (1):

V＝Vp+(y _i-y _0i)r (1)

6) the striped tube fluorescent screen luminosity of establishing current fringe number strong point correspondence is L _i, L then _iComputing formula be formula (2):

L _i＝Aj _i(V _i-V _d) ⁿ (2)

V in the formula _iFor current fringe number strong point correspondence this moment the striped pipe the accelerating potential value, A with n is and the relevant parameter of striped tube fluorescent screen medium, j _iElectron stream density for current fringe number strong point correspondence;

7) if under non-deflection situation, then current fringe number strong point is dropped on the static demarcating line, and remembers that the striped tube fluorescent screen brightness at current fringe number strong point is L _0i, L then _0iShown in (3) formula:

L _0i＝Aj _i(V _p-V _d) ⁿ (3)

8) can obtain the striped tube fluorescent screen brightness value L of current fringe number strong point on the static demarcating line by above (1) (2) (3) formula _0iBe formula (4), and hypothesis n=1:

$L_{0i}=L_i\×\left(\frac{V_p-V_d}{V_p-V_d+(y_i-y_{0i})r}\right)---\left(4\right)$

Then the weights coefficient of weighted correction is

9) pending image M _d' in (x, y _i) point gray values of pixel points G _XyiExpression;

10) with the ordinate y of static demarcating line ₀₁, y ₀₂... y _0i... y ₀₈For benchmark is treated processed image M respectively _d' go up the gray values of pixel points G of every stripe image _XyiMake weighted, weights are

Gray values of pixel points G after the conversion _Xyi' as the formula (5):

${G_{x\mathrm{yi}}}^{,}=G_{\mathrm{xyi}}\·\frac{V_p-V_d}{V_p-V_d+(y_i-y_{0i})r}---\left(5\right)$

Current as can be seen fringe number strong point is far away more apart from the static demarcating line, and the suffered accelerating potential in promptly current fringe number strong point is big more, and the be weakened the more powerful intensity image distortion at then current fringe number strong point of the intensity level at current fringe number strong point obtains proofreading and correct.

Beneficial effect

The present invention is to having carried out the weighted correction because the screen brightness that imaging mechanism causes distorts, method is by existing essential data, and operand is little, and is simple, and be common to laser radar system, especially real-time data handling system based on the three-dimensional imaging of striped pipe; Not only can reflect target property comparatively really, more follow-up range information provides accurate data.

Description of drawings

Fig. 1 is the dynamic image M of target _d

Fig. 2 is the dynamic image M after the process SNN filtering _D0

Fig. 3 is M for demarcating the pending image of finishing _d';

Fig. 4 is for proofreading and correct the image after finishing.

Embodiment

The present invention will be further described below in conjunction with drawings and Examples.

Embodiment

The bearing calibration of a kind of striped pipe imaging laser radar intensity image distortion, its concrete steps are:

1) still image of target is 400*400, to the still image M of target _sAdopt least square method to carry out static datum line and demarcate, obtain the ordinate y of the static demarcating line of single pixel wide ₀₁, y ₀₂... y _0i... y ₀₈The ordinate of the datum line that extracts is as shown in the table:

Striped

?y 01

y 02

y 03

y 04

y 05

y 06

y 07

y 08

Coordinate

?50

86

125

162

197

232

270

290

2) dynamic image of target is 400*400, as shown in Figure 1, and to the dynamic image M of target _dCarry out the SNN border and keep the class smothing filtering, its stencil value is 3; SNN is neighbour's mean filter, and the image that obtains is designated as M _D0, as shown in Figure 2;

3) according to the ordinate y of the static demarcating line that obtains ₀₁, y ₀₂... y _0i... y ₀₈At filtered dynamic image M _D0The corresponding surely straight line of last corresponding coordinate subscript, demarcating the pending image of finishing is M _d', as shown in Figure 3;

4) distance of measuring target as requested is 50m, sets the predeterminated voltage value V of striped pipe _pBe 1000V, the luminous required minimum voltage value V of striped tube fluorescent screen in the characteristic parameter of striped pipe _dBe 500V;

5) a semi-transparent semi-reflecting glass column is added the magnitude of voltage that scanning voltage is imaged as the point of a series of even distances and notes this series of points, the magnitude of voltage between any 2 of these a series of dibblings and the ratio between the distance value value that obtains of averaging is designated as r=0.68V;

If x is that current fringe number strong point is in pending image M _d' in horizontal ordinate, y _iFor current fringe number strong point in pending image M _d' in ordinate, the scanning voltage value V of the striped pipe of then current fringe number strong point correspondence is:

V＝1000+0.68(y _i-y _0i)

6) establish the striped tube fluorescent screen luminosity L of current fringe number strong point correspondence _iFor:

L _i＝Aj _i(V _i-500)

7) if under non-deflection situation, then current fringe number strong point is dropped on the static demarcating line, and remembers the striped tube fluorescent screen brightness L at current fringe number strong point _0iFor:

L _0i＝500Aj _i

8) according to step 5), 6) and 7) the striped tube fluorescent screen brightness value L of current fringe number strong point on the static demarcating line can be obtained _0iFor:

$L_{0i}=L_i\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68}$

The weights coefficient that is weighted correction is:

9) pending image M _d' in (x, y _i) point gray values of pixel points G _XyiExpression;

10) with the ordinate y of static demarcating line ₀₁, y ₀₂... y _0i... y ₀₈For benchmark is treated processed image M respectively _d' go up the gray values of pixel points G of every stripe image _XyiMake weighted, weights are

Gray values of pixel points G after the conversion _Xyi' be:

${G_{\mathrm{xyi}}}^{,}=G_{\mathrm{xyi}}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68}$

Promptly ${G_{\mathrm{xy}1}}^{,}=G_{\mathrm{xy}1}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₁, y _0i=y ₀₁=50;

${G_{\mathrm{xy}2}}^{,}=G_{\mathrm{xy}2}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₂, y _0i=y ₀₂=86;

${G_{\mathrm{xy}3}}^{,}=G_{\mathrm{xy}3}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₃, y _0i=y ₀₃=125;

${G_{\mathrm{xy}4}}^{,}=G_{\mathrm{xy}4}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₄, y _0i=y ₀₄=162;

${G_{\mathrm{xy}5}}^{,}=G_{\mathrm{xy}5}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₅, y _0i=y ₀₅=197;

${G_{\mathrm{xy}6}}^{,}=G_{\mathrm{xy}6}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₆, y _0i=y ₀₆=232;

${G_{\mathrm{xy}7}}^{,}=G_{\mathrm{xy}7}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₇, y _0i=y ₀₇=270;

${G_{\mathrm{xy}8}}^{,}=G_{\mathrm{xy}8}\·\frac{1000-500}{1000-500+(y_i-y_{0i})\×0.68},$ Wherein, y _i=y ₈, y _0i=y ₀₈=290;

With the Matlab software for calculation according to the above-mentioned processed image M for the treatment of respectively _d' in each pixel in every stripe bring above-mentioned formula into and travel through calculating, the image that obtains as shown in Figure 4, compare Fig. 3 as can be seen by Fig. 4, every stripe image is that benchmark has all obtained weakening in various degree with the static demarcating line, distortion in images has obtained correction, for follow-up intensity provides reliable accurate data as reconstruct and range information.

Claims (2)

1. the bearing calibration of striped pipe imaging laser radar intensity image distortion is characterized in that concrete steps are:

1) to the still image M of target _sAdopt least square method to carry out static datum line and demarcate, obtain the ordinate y of the i bar static demarcating line of single pixel wide _0i, all ordinates of static demarcating line are y ₀₁, y ₀₂... y _0i... y _0m, wherein i is 1～m, m is the fringe number of stripe pattern;

2) to the dynamic image M of target _dCarry out the border and keep the class smothing filtering, the image that obtains is designated as M _D0

3) according to the ordinate y of the static demarcating line that obtains ₀₁, y ₀₂... y _0i... y _0mAt filtered dynamic image M _D0The corresponding surely straight line of last corresponding coordinate subscript, demarcating the pending image of finishing is M _d';

4) measure the difference of target range distance as requested, set the predeterminated voltage value V of striped pipe _p, the luminous required minimum voltage value of striped tube fluorescent screen is V in the characteristic parameter of striped pipe _d

5) a semi-transparent semi-reflecting glass column is added the magnitude of voltage that scanning voltage is imaged as the point of a series of even distances and notes this series of points, the magnitude of voltage between any 2 of these a series of dibblings and the ratio between the distance value value that obtains of averaging is designated as r;

If x is that current fringe number strong point is in pending image M _d' in horizontal ordinate, y _iFor current fringe number strong point in pending image M _d' in ordinate, the scanning voltage value of the striped pipe of then current fringe number strong point correspondence is formula (1):

V＝V _p+(y _i-y _0i)r (1)

6) the striped tube fluorescent screen luminosity of establishing current fringe number strong point correspondence is L _i, L then _iComputing formula be formula (2):

L _i＝Aj _i(V _i-V _d) ⁿ (2)

V in the formula _iFor current fringe number strong point correspondence this moment the striped pipe the accelerating potential value, A with n is and the relevant parameter of striped tube fluorescent screen medium, j _iElectron stream density for current fringe number strong point correspondence;

7) if under non-deflection situation, then current fringe number strong point is dropped on the static demarcating line, and remembers that the striped tube fluorescent screen brightness at current fringe number strong point is L _0i, L then _0iShown in (3) formula:

L _0i＝Aj _i(V _p-V _d) ⁿ (3)

8) can obtain the striped tube fluorescent screen brightness value L of current fringe number strong point on the static demarcating line by above (1) (2) (3) formula _0iBe formula (4), and hypothesis n=1:

$L_{0i}=L_i\×\left(\frac{V_p-V_d}{V_p-V_d+(y_i-y_{0i})r}\right)---\left(4\right)$

Then the weights coefficient of weighted correction is

9) pending image M _d' in (x, y _i) point gray values of pixel points G _XyiExpression;

10) with the ordinate y of static demarcating line ₀₁, y ₀₂... y _0i... y ₀₈For benchmark is treated processed image M respectively _d' go up the gray values of pixel points G of every stripe image _XyiMake weighted, weights are

Gray values of pixel points G after the conversion _XyiAs the formula (5):

${G_{x\mathrm{yi}}}^{,}=G_{\mathrm{xyi}}\·\frac{V_p-V_d}{V_p-V_d+(y_i-y_{0i})r}---\left(5\right).$

2. the bearing calibration of a kind of striped pipe imaging laser radar intensity according to claim 1 image distortion is characterized in that: step 2) in to the dynamic image M of target _dCarrying out border maintenance class smothing filtering is neighbour's mean filter SNN, and its stencil value is 3.

Images