CN105091744A - Pose detection apparatus and method based on visual sensor and laser range finder - Google Patents

Abstract

The invention discloses a pose detection apparatus and method based on a visual sensor and laser range finders. One visual sensor and three laser range finders are utilized by the apparatus to together acquire information of a target. The method includes the steps: selecting from a target image characteristic points and characteristic lines, obtaining image coordinates of the characteristic points and angles of characteristic lines, and then obtaining image deviation of the characteristic points and angle deviation of the characteristic lines; calculating a current pixel equivalent, and obtaining deviation of a current pose in three degrees of freedom according to the current image deviation, the angle deviation and the pixel equivalent; and obtaining the angle deviation of the current pose in the corresponding two rotational degrees of freedom and the deviation of the current pose in a depth direction according to reading of the three laser range finders and relative position relations. In this way, six degrees of freedom of pose detection of a target can be realized.

Description

The apparatus for detecting position and posture of a kind of view-based access control model sensor and laser range finder and method

Technical field

The invention belongs to sensor detecting field, relate more specifically to apparatus for detecting position and posture and the method for a kind of view-based access control model sensor and laser range finder.

Background technology

Pose detects as important step, is widely used in the numerous areas such as robot controlling, aviation docking, homing guidance.Kind of sensor enriches, diverse in function, is the important tool obtaining external environment information.The information realization pose provided by sensor is detected and has attracted the concern of numerous scholar.

Vision sensor can the abundant environmental information of perception, and easy to install, is the comparatively normal a kind of external sensor used.Processed by the image information obtained vision sensor, the profile of target, shape, color can be obtained, the motion detection, relative positioning etc. of all right realize target.It is conventional position and posture detection method that the image information provided according to vision sensor obtains object pose information, these class methods are (such as with reference to SangJooKwon, HaemmJeong, andJaewoongHwang.KalmanFilter-BasedCoarse-to-FineControl forDisplayVisualAlignmentSystems [J] .IEEETransactionsonAutomationScienceandEngineering, 2012,9 (3): 621-628; BiaoZhang, JianjunWang, GregoryRossano, CarlosMartinezand kock.Vision-guidedRobotAlignmentforScalable, FlexibleAssemblyAutomation [C] .IEEEInternationalConferenceonRoboticsandBiomimetic.Phuk et, Thailand, 2011:944-951.) general based on monocular vision or binocular vision, first from visual pattern, extract key feature, then in conjunction with vision measurement technology, through multiple coordinate transform, obtain position and the attitude information of target, the pose of realize target detects.These class methods only use vision sensor, and sensor configuration is convenient, and testing process is simply efficient.But these class methods require that target is always completely visible in testing process, and the pose being therefore only applicable to small size target detects.

When not obtaining that abundant target information is as comparatively large in target size, partial visual feature is invisible, just need to complete target detection by other sensors.At present, conventional method mainly contains by laser tracker (for example, see ZhiLiu, YingXie, JingXu, KenChen.Lasertrackerbasedroboticassemblysystemforlargesc alepeg-holeparts [C] .IEEEInternationalConferenceonCyberTechnologyinAutomatio n, ControlandIntelligentSystems, HongKong, China, 2014:574-578.) or laser range finder (for example, see Young-KeunKim, YonghunKim, Kyung-sooKim, YunSubJung.DevelopingaRobustSensingSystemforRemoteRelati ve6-DOFMotionUsing1-DLaserSensors [C] .IEEEInternationalSystemsConference, Vancouver, CANADA.2012.) pose realizing large scale target detects.Laser tracker is suitable for the pose measurement of large scale parts very much, but expensive.Laser range finder measuring accuracy is high, stable performance, and antijamming capability is strong, and volume is little, is convenient to install.Vision sensor is arranged on target object by current method of carrying out pose detection by laser range finder and vision sensor, the pose of first manually adjustment laser range finder is needed during detection, it is made to get to the assigned address of target object, this class methods testing process is loaded down with trivial details, and accuracy of detection is lower.

Summary of the invention

Based on above-mentioned background, fundamental purpose of the present invention is the apparatus for detecting position and posture and the method that provide a kind of comprehensive visual sensor and laser range finder advantage.

For achieving the above object, as one aspect of the present invention, the invention provides the position and posture detection method of a kind of view-based access control model sensor and laser range sensor, comprising step as follows:

Step S0, uses a vision sensor and three laser range finders jointly to obtain target information;

Step S1: unique point and the characteristic straight line of selecting vision sensor sensitivity from the target image that described vision sensor obtains, off-line obtains unique point desired image coordinate and characteristic straight line expected angle, unique point described in On-line testing and characteristic straight line, obtain described unique point present image coordinate and described characteristic straight line current angular, relatively currency and expectation value, obtain the image deviations between current signature point and desired character point, and the deviation between current signature straight line angle and desired character straight line angle;

Step S2: according to the characteristic sum object space size of the described target image of On-line testing, obtain current pixel equivalent;

Step S3: according to described image deviations and current pixel equivalent, obtains current pose and the object pose deviation at the three degree of freedom of vision sensor sensitivity;

Step S4: according to reading and their relative position relation of described three laser range finders, off-line obtains the normal vector expectation value of the plane that described three laser range finders are formed, obtain current plane normal vector online, current normal vector and expectation normal vector are projected in plane corresponding to insensitive two rotary freedoms of vision respectively, and according to projecting to the component of each plane, obtaining current pose and expecting the angular deviation of attitude in corresponding degree of freedom;

Step S5: obtain current pose and the expected pose deviation at depth direction according to the reading of described three laser range finders;

Step S6: the deviation combining the three degree of freedom obtained from described target image and the deviation of three degree of freedom obtained from laser range finder reading, and current pose, obtain the six-degree-of-freedom information of target, the pose of realize target detects.

Wherein, described vision sensor is arranged on the position being convenient to obtain target object image, and described three laser range finders are isosceles triangle arrangement.

Wherein, the present image deviation of unique point described in step S1 and the current angular deviation of characteristic straight line are calculated by following formula:

$\left\{\begin{array}{c}\mathrm{\Δu}=u_d-u\\ \mathrm{\Δv}=v_d-v\\ {\mathrm{\θ}}_z={\mathrm{\θ}}_d-\mathrm{\θ}\end{array}\right.;$

Wherein, Δ u, Δ v respectively representation feature point in the image coordinate deviation in U direction and V direction, θ _zthe angular deviation of representation feature straight line.(u _d, v _d) representation feature point desired image coordinate, (u, v) representation feature point present image coordinate, θ _drepresentation feature straight line expected angle, θ representation feature straight line current angular.

Wherein, the current pixel equivalent described in step S2 is calculated by following formula:

t _s＝S/S ₀；

Wherein, t _srepresent current pixel equivalent, S represents target size in the picture, S ₀represent the bulk of target.

Wherein, the current pose described in step S3 and object pose are expressed as follows in the deviation of the three degree of freedom of vision sensor sensitivity:

$\left[\begin{array}{c}\mathrm{\Δx}\\ \mathrm{\Δy}\\ {\mathrm{\Δ\θ}}_z\end{array}\right]=\left[\begin{array}{c}{t}_s\·\mathrm{\Δu}\\ t_s\·\mathrm{\Δv}\\ {\mathrm{\θ}}_z\end{array}\right];$

Wherein, Δ x represents current pose and the expected pose position deviation in X-direction, and Δ y represents current pose and the expected pose position deviation in X-direction, Δ θ _zrepresent that current pose and expected pose are in the anglec of rotation deviation around optical axis direction.

Wherein, the current pose described in step S4 and object pose are expressed as follows in the deviation of insensitive two degree of freedom of vision sensor:

$\left\{\begin{array}{c}\cos \left({\mathrm{\Δ\θ}}_x\right)=\frac{V_{\mathrm{qyoz}}{V}_{\mathrm{yoz}}}{\left|V_{\mathrm{qyoz}}\right|\·\left|V_{\mathrm{yoz}}\right|}\\ \cos \left({\mathrm{\Δ\θ}}_y\right)=\frac{V_{\mathrm{qzox}}{V}_{\mathrm{zox}}}{\left|V_{\mathrm{qzox}}\right|\·\left|V_{\mathrm{yoz}}\right|}\end{array}\right.$

Wherein, Δ θ _xrepresent that current pose and object pose are in the anglec of rotation deviation around X-direction, Δ θ _yshow that current pose and object pose are in the anglec of rotation deviation around Y direction.V _qyozand V _qzoxthe expectation normal vector V of laser plane _qrespectively in the projection of YOZ plane and ZOX plane, V _yoxand V _zoxthat laser plane normal vector V is respectively in the projection of YOZ plane and ZOX plane.

Wherein, the current pose described in step S5 and the position deviation of object pose on the insensitive depth direction of vision sensor are expressed as follows:

Δz＝(d ₁+d ₂+d ₃)/3；

Wherein, Δ z represents current pose and the object pose position deviation in Z-direction, d _i(i=1,2,3) represent the reading of three laser range finders.

Wherein, described object pose is expressed as follows:

$\left[\begin{array}{c}x\\ y\\ z\\ {\mathrm{\θ}}_x\\ {\mathrm{\θ}}_y\\ {\mathrm{\θ}}_z\end{array}\right]=\left[\begin{array}{c}{x}_0\\ y_0\\ z_0\\ {\mathrm{\θ}}_{x0}\\ {\mathrm{\θ}}_{y0}\\ {\mathrm{\θ}}_{z0}\end{array}\right]+\left[\begin{array}{c}\mathrm{\Δx}\\ \mathrm{\Δy}\\ \mathrm{\Δz}\\ {\mathrm{\Δ\θ}}_x\\ {\mathrm{\Δ\θ}}_y\\ {\mathrm{\Δ\θ}}_z\end{array}\right];$

Wherein, [x, y, z, θ _x, θ _y, θ _z] represent and detect the object pose obtained, [x by vision sensor and laser range finder ₀, y ₀, z ₀, θ _x0, θ _y0, θ _z0] represent current pose, [Δ x, Δ y, Δ z, Δ θ _x, Δ θ _y, Δ θ _z] represent deviation between current pose and object pose, [Δ x, Δ y, Δ θ _z] be the pose deviation obtained by visual pattern, vision sensor to the pose sensitive of this three degree of freedom, [Δ z, Δ θ _x, Δ θ _y] be the pose deviation obtained according to the measured value of laser range finder, the posture information outside the three degree of freedom being obtained vision sensor sensitivity by laser range finder.

As another aspect of the present invention, present invention also offers the apparatus for detecting position and posture of a kind of view-based access control model sensor and laser range finder, comprise a vision sensor, three laser range finders and control module, wherein said vision sensor is arranged on the position being convenient to obtain target object image, described three laser range finders are isosceles triangle arrangement, and

The input of vision sensor and laser range finder as described in the position and posture detection method that described control module performs view-based access control model sensor as described in claim 1 to 8 any one and laser range finder controls, and calculating the six-degree-of-freedom information of target, the pose of realize target detects.

Wherein, described vision sensor is AVTGC1600H, and laser range finder is CASTALPT50220S.

Known based on technique scheme, the present invention has following beneficial effect: the position and posture detection method of traditional view-based access control model image requires that target is always completely visible in testing process, the pose being only applicable to small size target detects, for not obtaining abundant target image information as scenes such as target size are comparatively large, partial visual feature is invisible, it is difficult to the accurate posture information detecting target.The present invention is directed to the scene that vision sensor can not obtain abundant target image, jointly target information is obtained by vision sensor and laser range finder, the posture information of the three degree of freedom of vision sensor sensitivity is obtained from visual pattern, the posture information remaining three degree of freedom is obtained from laser distance information, merge the testing result of two kinds of sensors, the pose of realize target detects.This invention under the scene that cannot obtain abundant target image, according to visual image information and laser distance information, can obtain the six-freedom degree pose of target.A large amount of data on probation also demonstrate validity of the present invention.As can be seen here, the present invention, by the vision sensor of meticulous layout and laser sensor, solves the problem being only difficult to obtain abundant target information by vision sensor; The information that view-based access control model sensor and laser sensor provide, the six-freedom degree pose achieving target detects; Accuracy of detection is high, stability and real-time good.

Accompanying drawing explanation

Fig. 1 is the sensor placement figure of the apparatus for detecting position and posture of view-based access control model sensor of the present invention and laser range finder, the square expression laser range finder of its Oxford gray, and Dark grey circle represents vision sensor;

Fig. 2 is the process flow diagram of the position and posture detection method of view-based access control model sensor of the present invention and laser range finder;

Fig. 3 is the testing result of the apparatus for detecting position and posture sensor of view-based access control model sensor of the present invention and laser range finder.

Embodiment

Be described in detail embodiments of the invention below in conjunction with accompanying drawing: the present embodiment is implemented under premised on technical solution of the present invention, in conjunction with detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.

The invention discloses apparatus for detecting position and posture and the method for a kind of view-based access control model sensor and laser range finder, this device uses multiple sensors of meticulous layout jointly to gather target information, the information that dissimilar sensor provides is processed respectively, and result is merged, the six-freedom degree pose of realize target detects.

More specifically, as a preferred embodiment of the present invention, as Fig. 1 illustrates the sensor placement figure of the apparatus for detecting position and posture of view-based access control model sensor of the present invention and laser range finder, employ four sensors in this device: a vision sensor and three laser range finders.Vision sensor is arranged on the position being convenient to obtain target object image, and three laser range finders are isosceles triangle arrangement.For square target, sensor stand is also square, and three laser range finders are arranged on the left side of square set, the right and following midpoint respectively, and vision sensor is arranged on the top of square set.Wherein, described vision sensor is such as AVTGC1600H, and laser range finder is such as CASTALPT50220S.

The process flow diagram of the position and posture detection method of view-based access control model sensor of the present invention and laser range finder as shown in Figure 2, in testing process, the image deviations of current signature and desired character is obtained from visual pattern, and then obtain current pose and the expected pose information at the three degree of freedom of visual acuity, current pose and the expected pose position deviation at depth direction is obtained from laser distance information, and the angular deviation of current normal vector and expectation normal vector, thus obtain current pose and the information of expected pose on the insensitive three degree of freedom of vision, merge visual pattern result and laser distance information result, the pose of realize target detects, the method comprises the following steps:

The first step: unique point and the characteristic straight line of selecting vision sensor sensitivity from target image, off-line obtains unique point desired image coordinate and characteristic straight line expected angle, On-line testing unique point and characteristic straight line, obtain unique point present image coordinate and characteristic straight line current angular, relatively currency and expectation value, obtain the image deviations between current signature point and desired character point, and the deviation between current signature straight line angle and desired character straight line angle;

Second step: according to characteristics of image and the object space size of On-line testing, obtain current pixel equivalent;

3rd step: according to present image deviation and pixel equivalent, obtains current pose and the object pose deviation at the three degree of freedom of vision sensor sensitivity;

4th step: according to reading and their relative position relation of three laser range finders, off-line obtains the normal vector expectation value of the plane that three laser range finders are formed, obtain current plane normal vector online, current normal vector and expectation normal vector are projected in plane corresponding to insensitive two rotary freedoms of vision respectively, and according to projecting to the component of each plane, obtaining current pose and expecting the angular deviation of attitude in corresponding degree of freedom;

5th step: obtain current pose and the expected pose deviation at depth direction according to the reading of three laser sensors;

6th step: the deviation combining the three degree of freedom obtained from visual pattern and the deviation of three degree of freedom obtained from laser distance information, and current pose, obtain the six-degree-of-freedom information of target, the pose of realize target detects.

The described first step, specific as follows:

According to the target image that vision sensor obtains, extract key feature, the current angular deviation obtaining the present image deviation of unique point and characteristic straight line is as follows:

$\left\{\begin{array}{c}\mathrm{\Δu}=u_d-u\\ \mathrm{\Δv}=v_d-v\\ {\mathrm{\θ}}_z={\mathrm{\θ}}_d-\mathrm{\θ}\end{array}\right.$

Wherein, Δ u, Δ v respectively representation feature point in the image coordinate deviation in U direction and V direction, θ _zthe angular deviation of representation feature straight line.(u _d, v _d) representation feature point desired image coordinate, (u, v) representation feature point present image coordinate.θ _drepresentation feature straight line expected angle, θ representation feature straight line current angular.

Described second step, specific as follows:

According to target size in the picture and the bulk of target, obtain current pixel equivalent as follows:

t _s＝S/S ₀

Wherein, t _srepresent current pixel equivalent, S represents target size in the picture, S ₀represent the bulk of target.

Described 3rd step, specific as follows:

The image deviations obtained according to the first step and angular deviation, and the pixel equivalent that second step obtains, obtain current pose and object pose is as follows in the deviation of the three degree of freedom of vision sensor sensitivity:

$\left[\begin{array}{c}\mathrm{\Δx}\\ \mathrm{\Δy}\\ {\mathrm{\Δ\θ}}_z\end{array}\right]=\left[\begin{array}{c}{t}_s\·\mathrm{\Δu}\\ t_s\·\mathrm{\Δv}\\ {\mathrm{\θ}}_z\end{array}\right]$

Wherein, Δ x represents current pose and the expected pose position deviation in X-direction, and Δ y represents current pose and the expected pose position deviation in X-direction, Δ θ _zrepresent that current pose and expected pose are in the anglec of rotation deviation around optical axis direction.

Described 4th step, specific as follows:

By laser sensor as particle, then these three laser sensors form a plane.Remember that three laser sensor particles are A, B, C, then the planar process vector V of three particle formations is as follows:

$V=\stackrel{\→}{\mathrm{AB}}\×\stackrel{\→}{\mathrm{AC}}$

According to the measured value of three laser range finders and the position relationship between them, off-line obtains the planar process vector V expected _q, then at the planar process vector V that line computation is current, by V and V _qproject to YOZ plane and ZOX plane, then the angle of two normal vectors between the projection of YOZ plane is current pose and expected pose in the angular deviation around X-direction, and the angle of two normal vectors between the projection of ZOX plane is current pose and expected pose in the angular deviation around Y direction.Therefore, angular deviation is obtained as follows:

$\left\{\begin{array}{c}\cos \left({\mathrm{\Δ\θ}}_x\right)=\frac{V_{\mathrm{qyoz}}{V}_{\mathrm{yoz}}}{\left|V_{\mathrm{qyoz}}\right|\·\left|V_{\mathrm{yoz}}\right|}\\ \cos \left({\mathrm{\Δ\θ}}_y\right)=\frac{V_{\mathrm{qzox}}{V}_{\mathrm{zox}}}{\left|V_{\mathrm{qzox}}\right|\·\left|V_{\mathrm{yoz}}\right|}\end{array}\right.$

Wherein, Δ θ _xrepresent that current pose and object pose are in the anglec of rotation deviation around X-direction, Δ θ _yrepresent that current pose and object pose are in the anglec of rotation deviation around Y direction.V _qyozand V _qzoxthe expectation normal vector V of laser plane _qrespectively in the projection of YOZ plane and ZOX plane, V _yoxand V _zoxthat laser plane normal vector V is respectively in the projection of YOZ plane and ZOX plane.

Described 5th step, specific as follows:

According to the measured value of three laser range finders, obtain current pose and the position deviation of object pose on the insensitive depth direction of vision sensor is as follows:

Δz＝(d ₁+d ₂+d ₃)/3

Wherein, Δ z represents current pose and the object pose position deviation in Z-direction, d _i(i=1,2,3) represent the measured value of three laser range finders.

Described 6th step, specific as follows:

It is characterized in that, described object pose is expressed as follows:

$\left[\begin{array}{c}x\\ y\\ z\\ {\mathrm{\θ}}_x\\ {\mathrm{\θ}}_y\\ {\mathrm{\θ}}_z\end{array}\right]=\left[\begin{array}{c}{x}_0\\ y_0\\ z_0\\ {\mathrm{\θ}}_{x0}\\ {\mathrm{\θ}}_{y0}\\ {\mathrm{\θ}}_{z0}\end{array}\right]+\left[\begin{array}{c}\mathrm{\Δx}\\ \mathrm{\Δy}\\ \mathrm{\Δz}\\ {\mathrm{\Δ\θ}}_x\\ {\mathrm{\Δ\θ}}_y\\ {\mathrm{\Δ\θ}}_z\end{array}\right]$

Wherein, [x, y, z, θ _x, θ _y, θ _z] represent and detect the object pose obtained, [x by vision sensor and laser range finder ₀, y ₀, z ₀, θ _x0, θ _y0, θ _z0] represent current pose, [Δ x, Δ y, Δ z, Δ θ _x, Δ θ _y, Δ θ _z] represent deviation between current pose and object pose, [Δ x, Δ y, Δ θ _z] be the pose deviation obtained by visual pattern, vision sensor to the pose sensitive of this three degree of freedom, [Δ z, Δ θ _x, Δ θ _y] be the pose deviation obtained according to the measured value of laser range finder, the posture information outside the three degree of freedom being obtained vision sensor sensitivity by laser range finder.

In order to verify method of the present invention, pose detection is carried out to target.Fig. 3 is sensor detection results of the present invention, and (a) is target image result, and (b) is the measured value of three laser range finders.As can be seen from Figure 3, described method can detect target image accurately, and three laser ranging apparatus measuring values are basically identical, and now the pose deviation of pick-up unit and target is eliminated substantially, according to the current pose of device, the posture information of target can be obtained.

Above-described specific embodiment; object of the present invention, technical scheme and beneficial effect are further described; be understood that; the foregoing is only specific embodiments of the invention; be not limited to the present invention; within the spirit and principles in the present invention all, any amendment made, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (10)

1. a position and posture detection method for view-based access control model sensor and laser range finder, comprises the following steps:

Step S0, uses a vision sensor and three laser range finders jointly to obtain target information;

Step S1: unique point and the characteristic straight line of selecting vision sensor sensitivity from the target image that described vision sensor obtains, off-line obtains unique point desired image coordinate and characteristic straight line expected angle, unique point described in On-line testing and characteristic straight line, obtain described unique point present image coordinate and described characteristic straight line current angular, relatively currency and expectation value, obtain the image deviations between current signature point and desired character point, and the deviation between current signature straight line angle and desired character straight line angle;

Step S2: according to the characteristic sum object space size of the described target image of On-line testing, obtain current pixel equivalent;

Step S3: according to described image deviations and current pixel equivalent, obtains current pose and the object pose deviation at the three degree of freedom of vision sensor sensitivity;

Step S4: according to reading and their relative position relation of described three laser range finders, off-line obtains the normal vector expectation value of the plane that described three laser range finders are formed, obtain current plane normal vector online, current normal vector and expectation normal vector are projected in plane corresponding to insensitive two rotary freedoms of vision respectively, and according to projecting to the component of each plane, obtaining current pose and expecting the angular deviation of attitude in corresponding degree of freedom;

Step S5: obtain current pose and the expected pose deviation at depth direction according to the reading of described three laser range finders;

Step S6: the deviation combining the three degree of freedom obtained from described target image and the deviation of three degree of freedom obtained from laser range finder reading, and current pose, obtain the six-degree-of-freedom information of target, the pose of realize target detects.

2. the position and posture detection method of view-based access control model sensor according to claim 1 and laser range finder, wherein said vision sensor is arranged on the position being convenient to obtain target object image, and described three laser range finders are isosceles triangle arrangement.

3. the position and posture detection method of view-based access control model sensor according to claim 1 and laser range finder, wherein the present image deviation of unique point described in step S1 and the current angular deviation of characteristic straight line are calculated by following formula:

$\left\{\begin{array}{c}\mathrm{\Δu}=u_d-u\\ \mathrm{\Δv}=v_d-v\\ {\mathrm{\θ}}_z={\mathrm{\θ}}_d-\mathrm{\θ}\end{array}\right.;$

Wherein, Δ u, Δ v respectively representation feature point in the image coordinate deviation in U direction and V direction, θ _zthe angular deviation of representation feature straight line.(u _d, v _d) representation feature point desired image coordinate, (u, v) representation feature point present image coordinate, θ _drepresentation feature straight line expected angle, θ representation feature straight line current angular.

4. the position and posture detection method of view-based access control model sensor according to claim 1 and laser range finder, the current pixel equivalent wherein described in step S2 is calculated by following formula:

t _s＝S/S ₀；

Wherein, t _srepresent current pixel equivalent, S represents target size in the picture, S ₀represent the bulk of target.

5. the position and posture detection method of view-based access control model sensor according to claim 1 and laser range finder, the current pose wherein described in step S3 and object pose are expressed as follows in the deviation of the three degree of freedom of vision sensor sensitivity:

$\left[\begin{array}{c}\mathrm{\Δx}\\ \mathrm{\Δy}\\ \mathrm{\Δ}{\mathrm{\θ}}_z\end{array}\right]=\left[\begin{array}{c}{t}_s\·\mathrm{\Δu}\\ t_s\·\mathrm{\Δv}\\ {\mathrm{\θ}}_z\end{array}\right];$

Wherein, Δ x represents current pose and the expected pose position deviation in X-direction, and Δ y represents current pose and the expected pose position deviation in X-direction, Δ θ _zrepresent that current pose and expected pose are in the anglec of rotation deviation around optical axis direction.

6. the position and posture detection method of view-based access control model sensor according to claim 1 and laser range finder, the current pose wherein described in step S4 and object pose are expressed as follows in the deviation of insensitive two degree of freedom of vision sensor:

$\left\{\begin{array}{c}\cos \left({\mathrm{\Δ\θ}}_x\right)=\frac{V_{\mathrm{qyoz}}{V}_{\mathrm{yoz}}}{\left|V_{\mathrm{qyoz}}\right|\·\left|V_{\mathrm{yoz}}\right|}\\ \cos \left({\mathrm{\Δ\θ}}_y\right)=\frac{V_{\mathrm{qzox}}{V}_{\mathrm{zox}}}{\left|V_{\mathrm{qzox}}\right|\·\left|V_{\mathrm{yoz}}\right|}\end{array}\right.$

Wherein, Δ θ _xrepresent that current pose and object pose are in the anglec of rotation deviation around X-direction, Δ θ _yshow that current pose and object pose are in the anglec of rotation deviation around Y direction.V _qyozand V _qzoxthe expectation normal vector V of laser plane _qrespectively in the projection of YOZ plane and ZOX plane, V _yoxand V _zoxthat laser plane normal vector V is respectively in the projection of YOZ plane and ZOX plane.

7. the position and posture detection method of view-based access control model sensor according to claim 1 and laser range finder, the current pose wherein described in step S5 and the position deviation of object pose on the insensitive depth direction of vision sensor are expressed as follows:

Δz＝(d ₁+d ₂+d ₃)/3；

Wherein, Δ z represents current pose and the object pose position deviation in Z-direction, d _i(i=1,2,3) represent the reading of three laser range finders.

8. the position and posture detection method of view-based access control model sensor according to claim 1 and laser range finder, wherein said object pose is expressed as follows:

$\left[\begin{array}{c}x\\ y\\ z\\ {\mathrm{\θ}}_x\\ {\mathrm{\θ}}_y\\ {\mathrm{\θ}}_z\end{array}\right]=\left[\begin{array}{c}{x}_0\\ y_0\\ z_0\\ {\mathrm{\θ}}_{x0}\\ {\mathrm{\θ}}_{y0}\\ {\mathrm{\θ}}_{z0}\end{array}\right]+\left[\begin{array}{c}\mathrm{\Δx}\\ \mathrm{\Δy}\\ \mathrm{\Δz}\\ {\mathrm{\Δ\θ}}_x\\ {\mathrm{\Δ\θ}}_y\\ \mathrm{\Δ}{\mathrm{\θ}}_z\end{array}\right];$

Wherein, [x, y, z, θ _x, θ _y, θ _z] represent and detect the object pose obtained, [x by vision sensor and laser range finder ₀, y ₀, z ₀, θ _x0, θ _y0, θ _z0] represent current pose, [Δ x, Δ y, Δ z, Δ θ _x, Δ θ _y, Δ θ _z] represent deviation between current pose and object pose, [Δ x, Δ y, Δ θ _z] be the pose deviation obtained by visual pattern, vision sensor to the pose sensitive of this three degree of freedom, [Δ z, Δ θ _x, Δ θ _y] be the pose deviation obtained according to the measured value of laser range finder, the posture information outside the three degree of freedom being obtained vision sensor sensitivity by laser range finder.

9. the apparatus for detecting position and posture of a view-based access control model sensor and laser range finder, comprise a vision sensor, three laser range finders and control module, wherein said vision sensor is arranged on the position being convenient to obtain target object image, described three laser range finders are isosceles triangle arrangement, and

The input of vision sensor and laser range finder as described in the position and posture detection method that described control module performs view-based access control model sensor as described in claim 1 to 8 any one and laser range finder controls, and calculating the six-degree-of-freedom information of target, the pose of realize target detects.

10. the apparatus for detecting position and posture of view-based access control model sensor according to claim 9 and laser range finder, wherein said vision sensor is AVTGC1600H, and laser range finder is CASTALPT50220S.