CN105800464A - Positioning method based on automatic lifting hook system - Google Patents

Abstract

The invention relates to a positioning method based on an automatic lifting hook system. The method is the positioning method based on the automatic lifting hook system combining ultrasonic ranging and machine vision. The automatic lifting hook system is designed, and therefore accurate positioning of a lifting hook positioning point and a load positioning point is achieved. Z-direction positioning is achieved through an ultrasonic ranging method. A sensor is connected with a lifting hook main controller, the lifting hook main controller sends out a ranging command, a timer obtains the return time of ultrasonic waves, and the system calculates the distance H of a lifting hook in the Z direction. A visual positioning method is adopted to achieve positioning in the X direction and the Y direction. A visual positioning algorithm is designed by using load information in a known scene, expected positional information for work of the lifting hook system serves as a reference road sign, a road sign database is built to serve as reference conditions, real-time images and video information are subject to object identification and feature matching treatment, relative positional information needed by the lifting hook system is obtained, and accurate positioning of the lifting hook in the X direction and the Y direction is achieved.

Description

A kind of localization method based on automatic hanging hook system

Technical field

The invention belongs to technical field of engineering machinery, be specifically related to a kind of localization method towards the application of crane gear automatic hanging hook system.

Background technology

Automatic hanging hook system, for completing automatic job task, is determined the relative position of load and suspension hook by suspension hook anchor point and the combination of suspension ring anchor point, and the location of suspension hook and suspension ring needs to meet certain positioning precision, and otherwise automatic hanging hook system cannot normal operation.GPS and cellular radio location, positioning precision respectively 5 meters and 150 meters is had under outdoor environment.Technology such as mainly adopting optics, A-GPS, infrared ray, ultrasound wave, RFID, WIFI is positioned under indoor environment.The positioning precision of RFID technique is relevant with the distributed quantity of read write line, and representative is SPOTON system and LANDMARC system.Company of Qihoo adopts the location mode that GPS, WiFi, base station and gravity sensitive combine, and researches and develops a intelligent watch and realizes indoor 20 meters, the location of outdoor 5 meters, the real time position of real-time positioning wearer.The application of automatic hanging hook system needs to reach higher positioning precision, and traditional location technology cannot realize.It is furnished with one block of Magnet bottom the electromagnetic type automatic hanging hook of CarlStahl company of Germany R & D design, is caught suspension ring or hoist cable position by sucking action.Owing to suspension hook and periphery are all ferromagnetic materials, the positioning capturing operation of severe jamming hoist cable position, position inaccurate.Being accurately positioned of hanger system in the vertical direction can be realized for automatic hanging hook system location requirement, laser and ultrasonic ranging method.Consider shortage cost performance and the practicality of laser sensor, and ultrasonic sensor cost performance is high and simple in construction.

Summary of the invention

Present invention aim to address that what how to be effectively improved crane automatic hanging hook is accurately positioned problem, it is provided that the localization method of a kind of Engineering Oriented plant equipment automatic hanging hook system application.

Automatic hanging hook alignment system target is to realize being accurately positioned between suspension hook anchor point and load orientation point, Unify legislation under O-XYZ coordinate system.Wherein, X-direction is the direction that hanger system follows that crane horizontal cross moves, and Y-direction is the direction that hanger system follows that crane horizontally vertically moves, and Z-direction is the direction that lifting rope vertically moves；

Technical solution of the present invention

A kind of localization method based on automatic hanging hook system, the localization method that the method combines with machine vision based on ultrasonic ranging, designs automatic hanging hook system positioning method, it is achieved suspension hook anchor point 1 is accurately positioned with load orientation point 2.

Specifically comprising the following steps that of the method

1st, the connection circuit of ultrasonic sensor with suspension hook master controller composition is arranged on suspension hook anchor point, adopts ultrasonic ranging method to realize Z-direction location；Ultrasonic sensor is connected with suspension hook master controller starts working, and controller sends range finding order, and intervalometer obtains hyperacoustic time of return, the distance H in system-computed suspension hook Z-direction；

2nd, video camera and operating circuit are arranged on suspension hook anchor point, based on reference road sign (3) and scene load (4), adopt vision positioning method to realize X-direction and Y-direction location；Utilize known scene load information design vision localization algorithm, the desired locations information of automatic hanging hook system work is as reference road sign, set up landmark database as reference conditions, real time imaging and video information process through target recognition, characteristic matching, obtain the relative position information that hanger system needs, it is achieved suspension hook being accurately positioned in the x-direction and the z-direction.

Specifically comprising the following steps that of described vision positioning method

2.1st, the acquisition process of picture signal；Initializing program parameter, K=0, K represents location algorithm number of run；Video camera gathers video image information according to actual needs, and sets up the picture material storehouse of automatic hanging hook system vision localization；Video camera shooting image F (0), F (1) ..., F (i)；F (0) is with reference to road sign initial pictures, represents optimum visual anchor point image；F (i) represents that suspension hook is at image captured by the i moment；

2.2nd, vision localization parameter detecting and assessment result；Software load image, is loaded into image processing operators by collection image and positional parameter processes in operator；When geometry and the threedimensional model of known reference object, application camera marking method estimates the homography matrix of automatic hanging hook System planes and the plane of delineation, construct the constraint equation of camera intrinsic parameter and homography matrix, utilize method of least square that camera interior and exterior parameter carries out linear solution and nonlinear optimization, estimate camera intrinsic parameter, outer parameter and lens distortion parameter, eliminate the impact of optical distortion in image；Perform image processing operators；Image procossing obtain outer parameter matrix t (x, y, z), obtain positional parameter t (0), t (1) ..., t (i-1)；T (i) represents the translation vector parameter matrix between F (i-1) and F (i)；Calculation of position errors error；Error e rror represents the framing point in i moment and with reference to the position error between road sign initial alignment picture point；Obtain the real time kinematics carrier element local relative position information relative to the hanger system specific object of reference of three-dimensional scenic desired locations；

$error=\sqrt{{(x-\hat{x})}^2+{(y-\hat{y})}^2+{(z-\hat{z})}^2};$

2.3rd, output result；Be automatically performed out hook smoothly according to suspension hook with close hook operation need effective orientation distance, carry out error threshold correction；When error e rror is less than predetermined threshold value, meet expectation requirement；Otherwise, according to relative position information, drive crane system, mobile lift hook position, until it reaches target, it is achieved suspension hook automation remote accurately controls, complete the vision localization of automatic hanging hook system.

Advantages of the present invention and good effect:

The present invention is based on actual demands of engineering, design one automatic hanging hook system positioning method.It is hooked into row by vision localization technology and supersonic sensing range finder module auxiliary hanging to be accurately positioned, it is ensured that the safe and reliable work of automatic hanging hook system, is no longer rely on suction device and obtains location information by traditional methods such as artificial observations.Utilize vision localization information realization suspension hook to be accurately positioned, improve work efficiency, undoubtedly to realizing suspension hook Unmanned operation, promote safety and accuracy, there is clear advantage and good effect.

Accompanying drawing explanation

Fig. 1 is automatic hanging hook system positioning method schematic diagram；

Fig. 2 is automatic hanging hook system vision positioning system block diagram；

Fig. 3 is automatic hanging hook system vision localization algorithm；

Fig. 4 is that angular coordinate extracts result；

Fig. 5 is vision localization testing result；

Fig. 6 is vision localization experiment suspension hook characteristics of motion figure；

Fig. 7 is vision localization experimental test procedures；

Fig. 8 is vision localization experiment X, Y-direction test result；Wherein, a is 10mm positioning result, and b is 20mm positioning result, and c is 30mm positioning result, and d is 40mm positioning result；

Fig. 9 is vision localization experiment Z-direction test result；

Figure 10 is vision localization experimental results；

Figure 11 is the localization method application of overhead crane control system；

In figure, 1 suspension hook anchor point, 2 load orientation points, 3 with reference to road sign, 4 scene loads.

Detailed description of the invention

Embodiment 1:

1. automatic hanging hook system positioning method principle design involved in the present invention is as shown in Figure 1.Its target is to realize being accurately positioned between suspension hook anchor point 1 and load orientation point 2.Suspension hook anchor point determines the relative position of load and suspension hook with the combination of suspension ring anchor point, and vision localization result meets certain precision, and otherwise automatic hanging hook system cannot normal operation.

Design automatic hanging hook system location hardware platform:

Based on the localization method that ultrasonic ranging combines with machine vision, adopt automatic hanging hook system positioning method, it is achieved suspension hook anchor point 1 is accurately positioned with load orientation point 2.The connection circuit installation site that ultrasonic sensor and suspension hook master controller form is positioned at suspension hook anchor point, adopts ultrasonic ranging method to realize Z-direction location；Based on reference road sign 3 and scene load 4, video camera and operating circuit installation site are positioned at suspension hook anchor point, and the reference road sign designed in advance is fixed on load orientation point lower end.Selecting ultrasonic sensor to be connected with suspension hook master controller, controller sends range finding order, and intervalometer obtains hyperacoustic time of return, the distance H in system-computed suspension hook Z-direction.Vision positioning method is adopted to realize X-direction and Y-direction location.Utilize known scene load information design vision localization algorithm, the desired locations information of automatic hanging hook system work is as reference road sign, set up landmark database as reference conditions, real time imaging and video information process through target recognition, characteristic matching, obtain the relative position information that hanger system needs, it is achieved suspension hook being accurately positioned in the x-direction and the z-direction.

2. automatic hanging hook system vision positioning system block diagram is as shown in Figure 2.When geometry and the threedimensional model of known reference object, method with reference to camera calibration, computer vision algorithms make is combined with suspension hook automatic control technology, the visual image signal obtained with video camera carries out real-time feedback, obtain outer parameter matrix, design automatic hanging hook system vision localization algorithm, it is thus achieved that real time kinematics carrier element is relative to the local relative pose of the hanger system specific object of reference of three-dimensional scenic desired locations, it is achieved suspension hook automation remote accurately controls.Automatic hanging hook system vision localization algorithm is as it is shown on figure 3, specifically comprise the following steps that

1st, the acquisition process of picture signal.Initializing program parameter, K=0, K represents location algorithm number of run；Video camera gathers video image information according to actual needs, and sets up the picture material storehouse of hanger system vision localization；Video camera shooting image F (0), F (1) ..., F (i)；F (0) is with reference to road sign initial pictures, represents optimum visual anchor point image；F (i) represents that suspension hook is at image captured by the i moment.For realizing the experiment effect of the best, use 10～20 width lossless compress pictures.The scene of design object load comprises gridiron pattern image information, and gridiron pattern image is often taken as typical image and is applied to camera calibration technology.Harris-SIFT method can reflect the notable structure of image, and coupling correctness is high, and database volume is little, and algorithm real-time is good, therefore adopts Harris-SIFT feature extraction operator detection X-comers.The target of corner detection operator is the gridiron pattern image of detection hanger system load, extracts angular coordinate, as shown in Figure 4, and sets up imaging model.Application Matlab computer vision workbox processes hanger system vision localization algorithm, obtains vision localization parameter, it is achieved the precise positioning of suspension hook.Before location algorithm program brings into operation, the actual measured value of the input target gridiron pattern length of side, then image characteristic point detection processes each width checkerboard image that Programmable detection is loaded into.If image repeats, smudgy or the deviation of extreme angle occurs, then window can be pointed out out of court, result display total number of images and being accepted and picture number out of court.Pane display checkerboard image, circle instruction detection corner location, detected image uses label to carry out labelling, it is possible to the state plateau of real-time monitored arbitrary image test point can indicate coordinate information, including direction in chessboard plane of initial point, x-axis and y-axis.

Angle point minor shifts in any direction, the gray scale of its neighborhood all can change a lot.Detective operators realizes method: first look for the extreme point in DOG as candidate feature point, then passes through sub-pixel interpolation and improves positioning precision, filters the point of mobile rim.Directly detecting angle point in graphical rule space, the extreme point of metric space represents the extreme point of original image relevant position, is reused by Harris operator, saves amount of calculation.Harris operator formula:

$\mathrm{\μ}\left(x,{\mathrm{\σ}}_I,{\mathrm{\σ}}_s\right)=g\left({\mathrm{\σ}}_I\right)\×\left[\begin{array}{cc}{L_x}^2\left(x,{\mathrm{\σ}}_s\right)& L_x{L}_y\left(x,{\mathrm{\σ}}_s\right)\\ L_x{L}_y\left(x,{\mathrm{\σ}}_s\right)& {L_y}^2\left(x,{\mathrm{\σ}}_s\right)\end{array}\right]---\left(1\right)$

Cornerness=D (μ (x, σ_I,σ_s))-αT²(μ(x,σ_I,σ_s))(2)

In Harris operator formula, cornerness is Corner Feature value, and u is metric space second-order matrix, σ_IFor integral scale, σ_sFor characteristic point yardstick, L_αFor gradient operator.Owing to the error of approximation of convolution smooth operation and DoG can cause that candidate feature point and angle point are misaligned, there is the error of certain pixel, to neighborhood image processor.Eigenvalue is calculated, it may be judged whether there is angle point, if being absent from, then filtering candidate point, otherwise generating feature description for each point in the region that radius near candidate point is 2 or 3.Screening has the SIFT feature point of Corner Feature, it is desirable to each characteristic point, close to angle point, omits other not remarkable characteristics.

2nd, vision localization parameter detecting and assessment result.

Software load image, is loaded into image processing operators by collection image and positional parameter processes in operator；When geometry and the threedimensional model of known reference object, application camera marking method estimates the homography matrix of hanger system plane and the plane of delineation, construct the constraint equation of camera intrinsic parameter and homography matrix, utilize method of least square that camera interior and exterior parameter carries out linear solution and nonlinear optimization, estimate camera intrinsic parameter, outer parameter and lens distortion parameter, eliminate the impact of optical distortion in image.The geometric properties information of vision localization parameter processing operator application known target hook load, obtains homography transformation matrix according to projective transformation.Pose parameter between design vision localization parameter processing operator extraction suspension hook anchor point, homography matrix represents the pose transformation relation between load and video camera, comprises rotation relationship and translation relation.

Covering load plane by fixing for black and white gridiron pattern plane, demarcationization processes, and utilizes the resolution process algorithm of homography matrix to obtain outer parameter matrix solution, namely obtains spin matrix and translation matrix, obtains suspension hook positional parameter.Hanger system load orientation point X_w=(X_w,Y_w,Z_w)^TTransform to imaging point m=(u, v)^T.As shown in Equation 3, matrix P=K [R, t] is called perspective projection matrix, α_uAnd α_vRepresenting focal length, matrix K is camera intrinsic parameter matrix, the obliquity factor of behalf pixel.

$Z_c\left[\begin{array}{c}u\\ v\\ 1\end{array}\right]=\left[\begin{array}{ccc}{\mathrm{\α}}_u& s& u_0\\ 0& {\mathrm{\α}}_v& v_0\\ 0& 0& 1\end{array}\right](R\left[\begin{array}{c}{X}_w\\ Y_w\\ Z_w\end{array}\right]+t)=K({\mathrm{RX}}_w+t)---\left(3\right)$

If λ=Z_c, picture point coordinateThe coordinate of world point M isFrom P³To P²Linear projective transformation imaging model be

Suspension hook alignment system as shown in Figure 1, gridiron pattern plane, at XOY plane, uses r_iRepresent each column vector of R.The projective transformation relation of the angle point corresponding diagram picture point of gridiron pattern plane is as shown in Equation 4.

$s\left[\begin{array}{c}u\\ v\\ 1\end{array}\right]=K\[\begin{array}{cccc}{r}_1& r_2& r_3& t\end{array}\]\left[\begin{array}{c}X\\ Y\\ 0\\ 1\end{array}\right]=K\[\begin{array}{ccc}{r}_1& r_2& t\end{array}\]\left[\begin{array}{c}X\\ Y\\ 1\end{array}\right]=H_1\left[\begin{array}{c}X\\ Y\\ 1\end{array}\right]---\left(4\right)$

$H_1=K\[\begin{array}{ccc}{r}_1& r_2& t\end{array}\]=\left[\begin{array}{ccc}{h}_{11}& h_{12}& h_{13}\\ h_{21}& h_{22}& h_{23}\\ h_{31}& h_{32}& h_{33}\end{array}\right]---\left(5\right)$

H₁For homography matrix, there are 8 unknown quantitys to wait to solve, after arrangement, obtain linear equation.

$\left[\begin{array}{cccccccc}X& Y& 1& 0& 0& 0& -uX& -uY\\ 0& 0& 0& X& Y& 1& -vX& -vY\end{array}\right]h=\left[\begin{array}{c}u\\ v\end{array}\right]---\left(6\right)$

H=[h₁₁h₁₂h₁₃h₂₁h₂₂h₂₃h₃₁h₃₂]^T(7)

Choose the corresponding point of more than four groups in two width gridiron pattern plane pictures, utilize least squares identification unknown parameter, solve h, after normalization, obtain H₁.Homography matrix H₁Being Intrinsic Matrix and the coupling of outer parameter matrix, extracting pose parameter needs first to solve Intrinsic Matrix, then solves out parameter matrix.

H₁=[h₁h₂h₃]=K [r₁r₂t](8)

According to internal two constraints: one, r₁,r₂Orthogonal；Its two, rotating vector modulus value is 1.The K that inside enters a match comprises α_u、α_v、u₀、v₀And s, it is necessary to three homography matrixes solve intrinsic parameter.

Mobile suspension hook anchor point and the relative position of load orientation point, collect at least three width images as vision localization material database, under internal constraints, obtain at least six relational expression, and enter a match in thus solving K.Then, out parameter spin matrix R and translation matrix t is solved.Error e rror represents the framing point in i moment and with reference to the position error between road sign initial alignment picture point, image procossing obtains outer parameter matrix t (x, y, z), image testing result is as it is shown in figure 5, obtain one group of outer parameter matrix, positional parameter t (0) can be obtained, t (1) ..., t (i-1).T (i) represents the translation vector parameter matrix between F (i-1) and F (i), it is thus achieved that real time kinematics carrier element is relative to the local relative position information of the hanger system specific object of reference of three-dimensional scenic desired locations.Calculation of position errors error；

$error=\sqrt{{(x-\hat{x})}^2+{(y-\hat{y})}^2+{(z-\hat{z})}^2}---\left(10\right)$

3rd, output result；Be automatically performed out hook smoothly according to suspension hook with close hook operation need effective orientation distance, carry out error threshold correction；When error e rror is less than predetermined threshold value, meet expectation requirement；Otherwise, according to relative position information, drive crane system, mobile lift hook position, until it reaches target, it is achieved suspension hook automation remote accurately controls, complete automatic hanging hook system vision localization.

1) X, Y-direction positioning experiment

Planning crane moves at XOY plane, and photographic head distance load is 302mm, and the position keeping Z-direction is constant.Positioning experiment is arranged: the vision localization experiment suspension hook characteristics of motion is shown in Fig. 6, and black color dots represents target object initial position point, and target object according to arrow direction indication, moves the distance of setting successively every time along X or Y-direction.Vision localization experimental test procedures is as shown in Figure 7；Fig. 8 is vision localization experiment X, and Y-direction test result, wherein, a is 10mm positioning result, and b is 20mm positioning result, and c is 30mm positioning result, and d is 40mm positioning result；

2) Z-direction positioning experiment

The focal distance f of video camera is 4mm, and the displaced posi of load is all the time outside the focus of video camera, and the moving orthogonal scaling in Z-direction is 2.5, and positioning result is as shown in Figure 9.

As shown in Figure 10, the mean error of experiment value and theoretical value is 1mm to vision localization experimental error result, and position error fluctuation range is within 5mm, and location algorithm precision meets automatic hanging hook systematic difference requirement.

3) experiment test

Experiment test vision localization algorithm controls systematic difference at actual crane automatic hanging hook.Automatic hanging hook experimental prototype is put in overhead crane experiment porch, and application experiment is as shown in figure 11.By vision localization result, it is judged that X-direction needs mobile 315mm, using the X=315mm expectation input position as overhead crane automatic hanging hook system, drive trolley movement absolute distance 315mm by trolley movement controller, it is achieved the precise positioning of suspension hook.

List of references

[1] Ma Songde, Zhang Zhengyou. computer vision-theory of computation and algorithm basis [M]. Beijing: Science Press, 2003.

[2] ZZhang.Aflexiblenewtechniqueforcameracalibration [J] .IEEETransactiononPatternAnalysisandMachineIntelligence. 2000,22 (11): 1330-1334.

[3] Yang Shanhu, Li Yanming, Liu Chengliang. based on the tower crane picture control [J] of suspension hook action intelligent-tracking. electromechanical integration .2010,5 (4): 25-28.

[4] RichardHartley, AndrewZisserman work. Wei Sui, Yang Shangjun, Zhang Quanbing, Hu Maolin translate. the multi-view geometry [M] in computer vision. publishing house of University of Anhui .2002.

[5] Meng Xianyao, Han Xinjie, white wide next. the control of operation on the sea multipupose vessel automatic control system-dynamic positioning system and redundancy [J]. Worldwide Shipping .2004,27 (3).

[6] Yang Zhida, Liu Shaoyou. Precise Orientation Technique for Auto-Control of Cranes for Nuclear Waste Storehouse [J]. Nuclear Power Engineering .2006,27 (6): 94～98.

[7] WEIGQ, MASD.Implicitandexlicitcameracalibration:Theoryandexperi ments [J] .IEEETransactionsonPatternAnalysisandMachineIntelligence .1994,16 (5): 469-480.

[8] Faugeras, D.&Toscani, G.Cameracalibrationfor3Dcomputervision.ProcIntemationalW orkshoponIndustrial [J] .ApplicationsofMachineVisionandMachineIntelligence.Silke n, Japan, 1987:240-247.

[9]J.HeikkilaandO.Silven.Afour-stepcameracalibrationprocedurewithimplicitimagecorrection[J].ComputerVisionandPatternRecognition.1997.。

Claims (2)

1. the localization method based on automatic hanging hook system, it is characterized in that, the localization method that the method combines with machine vision based on ultrasonic ranging, designs automatic hanging hook system positioning method, it is achieved suspension hook anchor point (1) is accurately positioned with load orientation point (2)；

Wherein, X-direction is the direction that hanger system follows that crane horizontal cross moves, and Y-direction is the direction that hanger system follows that crane horizontally vertically moves, and Z-direction is the direction that lifting rope vertically moves；

Specifically comprising the following steps that of the method

1st, the connection circuit of ultrasonic sensor with suspension hook master controller composition is arranged on suspension hook anchor point, adopts ultrasonic ranging method to realize Z-direction location；Ultrasonic sensor is connected with suspension hook master controller starts working, and controller sends range finding order, and intervalometer obtains hyperacoustic time of return, the distance H in system-computed suspension hook Z-direction；

2nd, video camera and operating circuit are arranged on suspension hook anchor point, based on reference road sign (3) and scene load (4), adopt vision positioning method to realize X-direction and Y-direction location；Utilize known scene (4) load information design vision localization algorithm, the desired locations information of automatic hanging hook system work is as reference road sign (3), set up landmark database as reference conditions, real time imaging and video information process through target recognition, characteristic matching, obtain the relative position information that hanger system needs, it is achieved suspension hook being accurately positioned in the x-direction and the z-direction.

2. localization method according to claim 1, it is characterised in that specifically comprising the following steps that of the vision positioning method described in the 2nd step

2.1st, the acquisition process of picture signal；Initializing program parameter, K=0, K represents location algorithm number of run；Video camera gathers video image information according to actual needs, and sets up the picture material storehouse of automatic hanging hook system vision localization；Video camera shooting image F (0), F (1) ..., F (i)；F (0) is with reference to road sign initial pictures, represents optimum visual anchor point image；F (i) represents that suspension hook is at image captured by the i moment；

2.2nd, vision localization parameter detecting and assessment result；Software load image, is loaded into image processing operators by collection image and positional parameter processes in operator；When geometry and the threedimensional model of known reference object, application camera marking method estimates the homography matrix of automatic hanging hook System planes and the plane of delineation, construct the constraint equation of camera intrinsic parameter and homography matrix, utilize method of least square that camera interior and exterior parameter carries out linear solution and nonlinear optimization, estimate camera intrinsic parameter, outer parameter and lens distortion parameter, eliminate the impact of optical distortion in image；Perform image processing operators；Image procossing obtain outer parameter matrix t (x, y, z), obtain positional parameter t (0), t (1) ..., t (i-1)；T (i) represents the translation vector parameter matrix between F (i-1) and F (i)；Calculation of position errors error；Error e rror represents the framing point in i moment and with reference to the position error between road sign initial alignment picture point；Obtain the real time kinematics carrier element local relative position information relative to the hanger system specific object of reference of three-dimensional scenic desired locations；

$\mathrm{error}=\sqrt{{(x-\hat{x})}^2+{(y-\hat{y})}^2+{(z-\hat{z})}^2};$

2.3rd, export result；Be automatically performed out hook smoothly according to suspension hook with close hook operation need effective orientation distance, carry out error threshold correction；When error e rror is less than predetermined threshold value, meet expectation requirement；Otherwise, according to relative position information, drive crane system, mobile lift hook position, until it reaches target, it is achieved suspension hook automation remote accurately controls, complete the vision localization of automatic hanging hook system.