A method, apparatus, device and readable storage medium for identifying a profile of a hoisted object

Abstract

The application discloses a method, device and equipment for identifying a hoisted object profile and a readable storage medium, and comprises the following steps: acquiring a static point cloud map representing ground information in a tower crane working area when the tower crane is not in use; acquiring a real-time point cloud map representing an area below a trolley of the tower crane when the tower crane is in use; comparing the static point cloud map and the real-time point cloud map, and forming a difference point cloud map by points only existing in the real-time point cloud map but not existing in the static point cloud map; clustering points representing a steel wire rope at an upper end of a hook, points representing the hook, points representing a hoisting rope and points representing a hoisted object from the difference point cloud map by using a preset clustering algorithm; and calculating profile information of the hoisted object according to position information of each point in the points representing the hoisted object. The application can automatically calculate the profile size of the hoisted object, and is efficient and accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image recognition, and in particular relates to a method and device for identifying the outline of a hoisted object, equipment and a readable storage medium. BACKGROUND

[0002] In the process of using a tower crane, even if the starting position and the ending position of a hoisting operation are known, a safe and effective operation path cannot be planned, and the outline of the hoisted object also needs to be known so as to plan a safe operation path that can avoid collision between the hoisted object and obstacles in the working area of the tower crane; in the prior art, the outline of the hoisted object needs to be obtained by the staff in advance and manually input into a terminal for planning the operation path, which is a relatively cumbersome process. Therefore, how to automatically obtain the outline of the hoisted object so as to facilitate subsequent path planning has become a technical problem that needs to be solved by those skilled in the art. SUMMARY

[0003] The present application aims to provide a method and device for identifying the outline of a hoisted object, equipment and a readable storage medium, which can automatically calculate the outline of the hoisted object, and is not only efficient but also has good accuracy.

[0004] According to one aspect of the present application, a method for identifying the outline of a hoisted object is provided, which comprises:

[0005] obtaining a static point cloud map representing the ground information in the working area of the tower crane when the tower crane is not in use;

[0006] obtaining a real-time point cloud map representing the area below the trolley of the tower crane when the tower crane is in use;

[0007] comparing the static point cloud map and the real-time point cloud map, and forming a difference point cloud map from the points that exist only in the real-time point cloud map but not in the static point cloud map;

[0008] using a preset clustering algorithm to cluster the points representing the steel wire rope at the upper end of the hook, the points representing the hook, the points representing the hoisting rope, and the points representing the hoisted object from the difference point cloud map;

[0009] calculating the outline information of the hoisted object according to the position information of each point in the points representing the hoisted object.

[0010] Optionally, the static point cloud map, the real-time point cloud map and the difference point cloud map all adopt a preset tower crane coordinate system; wherein the tower crane coordinate system is established with the intersection of the tower crane and the ground as the origin, the positive east direction as the X axis, the positive north direction as the Y axis, and the upward direction perpendicular to the ground as the Z axis.

[0011] Optionally, the comparing the static point cloud map and the real-time point cloud map forms a difference point cloud map with points only existing in the real-time point cloud map but not in the static point cloud map, comprising:

[0012] Obtaining a current coordinate position of the tower crane trolley in the tower crane coordinate system;

[0013] Taking the current coordinate position as a positioning point, a reference point cloud map is intercepted from the static point cloud map; wherein, a ground area represented by the reference point cloud map is the same as a ground area represented by the real-time point cloud map;

[0014] Comparing the reference point cloud map and the real-time point cloud map, points only existing in the real-time point cloud map but not in the reference point cloud map form the difference point cloud map.

[0015] Optionally, the taking the current coordinate position as the positioning point, the reference point cloud map is intercepted from the static point cloud map, comprising:

[0016] Setting X-axis coordinate and Y-axis coordinate in the current coordinate position as reference points, and setting a reference radius according to a size of a ground area corresponding to the real-time point cloud map;

[0017] Forming a first set with X-axis coordinate and Y-axis coordinate of each point in the static point cloud map, and creating an empty second set;

[0018] Iterating each point in the first set in turn, and judging whether a distance between the current iterated point and the reference point is less than the reference radius;

[0019] If yes, the current iterated point is added to the second set, and if no, the next point in the first set is iterated;

[0020] When all points in the first set are iterated, the reference point cloud map is formed according to all points in the second set.

[0021] Optionally, the comparing the reference point cloud map and the real-time point cloud map forms the difference point cloud map with points only existing in the real-time point cloud map but not in the reference point cloud map, comprising:

[0022] Inputting the reference point cloud map and the real-time point cloud map into an iterative closest point (ICP) algorithm to obtain a rotation parameter R* and a translation parameter t* used for aligning and adjusting the real-time point cloud map;

[0023] Aligning and adjusting the real-time point cloud map according to the following formula to obtain a standard real-time point cloud map:

[0024] Standard real-time point cloud map = Real-time point cloud map * R * +t * ;

[0025] Input the standard real-time point cloud map, the reference point cloud map and a preset distance threshold into a K-Dimensional tree matching algorithm to determine the difference point cloud map.

[0026] Optionally, the clustering of the point cloud representing the upper end of the steel wire rope of the hook, the point cloud representing the hook, the point cloud representing the hoisting rope and the point cloud representing the hoisted object from the difference point cloud map by using a preset clustering algorithm comprises:

[0027] Obtaining a clustering model trained in advance based on a Euclidean clustering algorithm;

[0028] Inputting the three-dimensional coordinate position of each point in the difference point cloud map in the tower crane coordinate system into the clustering model to obtain the point cloud representing the upper end of the steel wire rope of the hook, the point cloud representing the hook, the point cloud representing the hoisting rope and the point cloud representing the hoisted object.

[0029] Optionally, the calculation of the contour information of the hoisted object according to the position information of each point in the point cloud representing the hoisted object comprises:

[0030] Obtaining the three-dimensional coordinate position of each point in the point cloud representing the hoisted object in the tower crane coordinate system, and taking the difference between the maximum Z-axis coordinate and the minimum Z-axis coordinate as the height of the hoisted object;

[0031] Obtaining a two-dimensional image according to the X-axis coordinate and the Y-axis coordinate of each point in the point cloud representing the hoisted object;

[0032] Dividing the two-dimensional image into N*N blocks according to a preset value N, and calculating the main direction angle of each block;

[0033] Calculating the main direction angle of the two-dimensional image according to the main direction angle of all blocks;

[0034] Rotating the two-dimensional image according to the main direction angle of the two-dimensional image to obtain a standard two-dimensional image;

[0035] Calculating the length and the width of the hoisted object according to the point located at the upper left corner and the point located at the lower right corner in the standard two-dimensional image.

[0036] In order to achieve the above-mentioned purpose, the application further provides a device for identifying the contour of a hoisted object, which comprises:

[0037] A first obtaining module is configured to obtain a static point cloud map representing the ground information in the working area of a tower crane when the tower crane is not in use;

[0038] A second acquisition module is configured to acquire a real-time point cloud map representing an area below the trolley of the tower crane when the tower crane is in use.

[0039] A point cloud comparison module is configured to compare the static point cloud map and the real-time point cloud map, and form a difference point cloud map from points only existing in the real-time point cloud map but not existing in the static point cloud map.

[0040] A point cloud clustering module is configured to cluster, from the difference point cloud map, point clouds representing the upper end of the steel wire rope of the hook, point clouds representing the hook, point clouds representing the hoisting rope, and point clouds representing the hoisted object, by using a preset clustering algorithm.

[0041] A contour identification module is configured to calculate contour information of the hoisted object according to position information of each point in the point clouds representing the hoisted object.

[0042] To achieve the above object, the present application further provides a computer device, which specifically comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above-mentioned method for identifying the contour of the hoisted object when executing the computer program.

[0043] To achieve the above object, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program implements the steps of the above-mentioned method for identifying the contour of the hoisted object when executed by a processor.

[0044] The method, device, equipment and readable storage medium for identifying the contour of the hoisted object provided by the present application can automatically calculate the contour size of the hoisted object by first acquiring a static point cloud map and a real-time point cloud map, then determining a difference point cloud map not existing in the static point cloud map from the real-time point cloud map, then identifying the hook, the hoisting rope and the hoisted object from the difference point cloud map, and finally calculating the length, width and height of the hoisted object based on the position information of the point clouds representing the hoisted object, compared with the prior art. BRIEF DESCRIPTION OF DRAWINGS

[0045] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of the preferred embodiments and are not intended to limit the scope of the application. Moreover, the same reference numerals in different figures represent the same or similar components. In the drawings:

[0046] Figure 1 An optional flowchart of the method for identifying the contour of the hoisted object provided by Example One;

[0047] Figure 2A schematic diagram of a tower crane provided for Embodiment One;

[0048] Figure 3 A schematic diagram of a static point cloud map provided for Embodiment One;

[0049] Figure 4 A schematic diagram of a real-time point cloud map provided for Embodiment One;

[0050] Figure 5 A schematic diagram of a clustering result based on a difference point cloud map provided for Embodiment One;

[0051] Figure 6 A schematic diagram of an initial two-dimensional image provided for Embodiment One;

[0052] Figure 7 A schematic diagram of a standard two-dimensional image provided for Embodiment One;

[0053] Figure 8 An alternative structural schematic diagram of a device for identifying a hoisted object profile provided for Embodiment Two;

[0054] Figure 9 An alternative hardware architecture schematic diagram of a computer device provided for Embodiment Three. DETAILED DESCRIPTION

[0055] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0056] Embodiment One

[0057] The embodiments of the present application provide a method for identifying a hoisted object profile, as shown in the accompanying drawings, which specifically comprises the following steps: Figure 1

[0058] Step S101: When the tower crane is not in use, a static point cloud map representing ground information in a working area of the tower crane is acquired.

[0059] Specifically, the static point cloud map can be acquired by a laser radar installed on a trolley of the tower crane; as shown in the accompanying drawings, Figure 2 ​As shown, a tower crane schematic diagram is shown, a tower crane trolley is arranged on the tower crane jib, and a laser radar, a trolley luffing ranging, and a trolley downward camera are arranged on the tower crane trolley. In addition, a wind sensor, a weight sensor, a rotation sensor, an inclination sensor, and a boom root camera are arranged at the connection between the tower crane jib and the tower crane tower body. In actual application, by rotating the tower crane jib and moving the tower crane trolley, the ground point cloud data in the entire working area of the tower crane can be collected by the laser radar, that is, a static point cloud map is obtained. In the static point cloud map, the information (position and height) of all ground obstacles is included. It should be noted that the ground information usually does not change in a period of time, so the static point cloud map can be obtained regularly according to the set time interval to reflect the ground information in a specific period of time.

[0060] Step S102: obtaining a real-time point cloud map representing the area below the tower crane trolley during use of the tower crane.

[0061] Specifically, the real-time point cloud map is obtained by the laser radar installed on the tower crane trolley. For example, Figure 3 As shown, the static point cloud map represents the entire working area of the tower crane, and the information of the hook, the hoisted object, and the hoisting rope is not included in the static point cloud map. As shown, Figure 4 As shown, the real-time point cloud map only represents the local area below the tower crane trolley, and in addition to the information of the ground obstacles, the information of the hook, the hoisted object, and the hoisting rope is also included in the real-time point cloud map.

[0062] Step S103: comparing the static point cloud map and the real-time point cloud map, and forming a difference point cloud map by points only existing in the real-time point cloud map but not existing in the static point cloud map.

[0063] As described above, the area corresponding to the real-time point cloud map is a small part of the area corresponding to the static point cloud map, and compared with the static point cloud map, the information of the hook, the hoisted object, and the hoisting rope is added in the real-time point cloud map. Therefore, by comparing the points in the two point cloud maps, the difference points only existing in the real-time point cloud map but not existing in the static point cloud map can be determined, and all the difference points determined are formed into the difference point cloud map.

[0064] Step S104: using a preset clustering algorithm to cluster the point cloud representing the upper end of the steel wire rope of the hook, the point cloud representing the hook, the point cloud representing the hoisting rope, and the point cloud representing the hoisted object from the difference point cloud map.

[0065] Specifically, a clustering algorithm for dividing four regions from the point cloud map can be trained in advance based on a large amount of sample data; preferably, the clustering algorithm can adopt a Euclidean clustering algorithm based on distance or a DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering algorithm based on density. As shown in FIG. 6, a clustering result is a schematic diagram for dividing the difference point cloud map into point clouds representing the steel wire rope, point clouds representing the hook, point clouds representing the sling, and point clouds representing the hoisted object. Figure Five

[0066] Step S105: calculating the contour information of the hoisted object according to the position information of each point in the point clouds representing the hoisted object; wherein the contour information includes length, width, and height.

[0067] Specifically, the static point cloud map, the real-time point cloud map, and the difference point cloud map all adopt a preset tower crane coordinate system; wherein the tower crane coordinate system is established with the intersection of the tower crane and the ground as the origin, the positive east direction as the X-axis, the positive north direction as the Y-axis, and the upward direction perpendicular to the ground as the Z-axis; in step S105, the position information of each point in the point clouds representing the hoisted object in the tower crane coordinate system is obtained, and the difference between the maximum coordinate value and the minimum coordinate value on the Z-axis is taken as the height of the hoisted object, the difference between the maximum coordinate value and the minimum coordinate value on the X-axis is taken as the length of the hoisted object, and the difference between the maximum coordinate value and the minimum coordinate value on the Y-axis is taken as the width of the hoisted object.

[0068] Through the above steps S101 to S105, the contour information of the hoisted object can be calculated using the static point cloud map and the real-time point cloud map collected by the laser radar on the tower crane; compared with the prior art, the staff does not need to obtain the contour information of the hoisted object in advance, and the contour size of the hoisted object can be automatically calculated by first obtaining the static point cloud map and the real-time point cloud map, then determining the difference point cloud map that does not exist in the static point cloud map from the real-time point cloud map, then identifying the hook, the sling, and the hoisted object from the difference point cloud map, and finally calculating the length, the width, and the height of the hoisted object based on the position information of the point clouds representing the hoisted object. The embodiment can automatically calculate the contour size of the hoisted object, which is not only efficient but also has good accuracy.

[0069] Further, when implementing the step of comparing the static point cloud map and the real-time point cloud map in step S103 to form the difference point cloud map only from the points existing in the real-time point cloud map but not in the static point cloud map, the step specifically includes:

[0070] Step A1: obtaining the current coordinate position of the tower crane trolley in the tower crane coordinate system;

[0071] ​Step A2: taking the current coordinate position as a positioning point, a reference point cloud map is intercepted from the static point cloud map; wherein, the ground area represented by the reference point cloud map is the same as the ground area represented by the real-time point cloud map;

[0072] Step A3: comparing the reference point cloud map and the real-time point cloud map, points only existing in the real-time point cloud map but not existing in the reference point cloud map are formed into the difference point cloud map.

[0073] In the manner of the above steps A1 to A3, the reference point cloud map corresponding to the implementation point cloud map is cut out from the static point cloud map, and the difference point cloud map is calculated based on the reference point cloud map, so that the operation amount of the comparison process is reduced, and the calculation efficiency is improved.

[0074] Further, the step A1 specifically comprises:

[0075] Step A11: obtaining an included angle γ between an initial position of the tower crane jib and the due east direction, and obtaining a rotation angle θ of the tower crane jib when the tower crane is used;

[0076] Step A12: obtaining a distance r of the tower crane trolley from the origin of the tower crane coordinate system in a top view, and obtaining a ground clearance h of the tower crane jib;

[0077] Step A13: calculating the current position (x, y, z) of the tower crane trolley in the tower crane coordinate system according to the following formula:

[0078] x = r*cos(γ+θ);

[0079] y = r*sin(γ+θ);

[0080] z = h.

[0081] In actual application, the polar coordinates (r, θ, h) of the current trolley in the tower crane coordinate system can be obtained through the values of the absolute encoders of the three major mechanisms of amplitude variation, rotation and lifting; for example, γ and θ can be obtained through the rotation sensor installed on the tower crane, and the distance r can be obtained through the trolley amplitude variation distance sensor.

[0082] Further, the step A2 specifically comprises:

[0083] Step A21: setting the X-axis coordinate and the Y-axis coordinate in the current coordinate position as reference points (x, y), and setting a reference radius R according to the size of the ground area corresponding to the real-time point cloud map;

[0084] It should be noted that the radius of the ground area corresponding to the real-time point cloud map is calculated according to the FOV (Field of View) of the laser radar and the height of the tower crane trolley from the ground, and the reference radius is slightly larger than the radius of the ground area corresponding to the real-time point cloud map.

[0085] Step A22: Form a first set of X-axis coordinates and Y-axis coordinates of each point in the static point cloud map, and create an empty second set;

[0086] Step A23: Traverse each point in the first set in turn, and determine whether the distance between the current traversed point and the reference point is less than the reference radius;

[0087] That is, calculate whether the distance d of the point (x j ′ , y j ′ ) in the first set to the reference point (x, y) is less than the reference radius R; wherein j∈[1, M], M is the total number of points in the first set;

[0088] Step A24: If yes, add the current traversed point to the second set, and if no, traverse the next point in the first set;

[0089] Step A25: When all points in the first set are traversed, form the reference point cloud map according to all points in the second set;

[0090] It should be noted that the points in the second set only include X-axis coordinates and Y-axis coordinates, and when forming the reference point cloud map, the Z-axis coordinates of each point in the second set also need to be added to form the reference point cloud map according to the three-dimensional coordinates of each point in the second set.

[0091] Further, the step A3 specifically comprises:

[0092] Step A31: Input the reference point cloud map and the real-time point cloud map into the iterative closest point (ICP) algorithm to obtain a rotation parameter R* and a translation parameter t* for aligning and adjusting the real-time point cloud map;

[0093] ICP (Iterative Closest Point) algorithm is a point cloud matching algorithm, the input of the ICP algorithm is the reference point cloud map and the real-time point cloud map, and the real-time point cloud map is taken as the source point cloud map, and the reference point cloud map is taken as the target point cloud map; the output of the ICP algorithm is the rotation parameter R* and the translation parameter t*, and the reference point cloud map and the real-time point cloud map are highly coincident by using the rotation parameter R* and the translation parameter t*; when the ICP algorithm is used, multiple iterations will be experienced, and each iteration will match n points in the real-time point cloud map from the reference point cloud map to obtain the corresponding n closest points; the pose parameter (R*, t*) obtained by using the following formula is used on the real-time point cloud map, and then the new pose parameter (R*, t*) is calculated again with the reference point cloud map, until the upper limit of the number of iterations or the final pose parameter is obtained after convergence; preferably, the specific use process of the ICP algorithm is as follows:

[0094] The centroid position p of the reference point cloud map is calculated according to the coordinate position of each point in the reference point cloud map, and the centroid position p' of the real-time point cloud map is calculated according to the coordinate position of each point in the real-time point cloud map;

[0095] The rotation parameter R* is calculated according to the following formula:

[0096]

[0097] Wherein, n is the total number of points in the reference point cloud map;

[0098] q i =p i -p, and p i is the i th point in the reference point cloud map;

[0099] q′ i =p′ i -p', and p' i is the i th point in the real-time point cloud map corresponding to the i th point in the reference point cloud map;

[0100] The translation parameter t* is calculated according to the following formula:

[0101] t * =p-R * p'.

[0102] Step A32: align and adjust the real-time point cloud map according to the following formula to obtain a standard real-time point cloud map:

[0103] Standard real-time point cloud map = real-time point cloud map × R * +t* ;

[0104] Step A33: inputting the standard real-time point cloud map, the reference point cloud map and a preset distance threshold into a K-Dimensional tree matching algorithm to determine the difference point cloud map.

[0105] In actual application, the weight of the hoisted object will cause the tilt of the tower crane jib, thereby causing the real-time point cloud map obtained by the laser radar installed on the tower crane jib to deviate by 1-5 meters, so it is necessary to correct the real-time point cloud map in the manner of steps A31 to A32 to obtain the corrected standard real-time point cloud map. In addition, the K-Dimensional tree matching algorithm, i.e. KD-Tree, is a tree-shaped data structure for storing instance points in K-dimensional space for fast retrieval. It is mainly applied to the search of key data in multi-dimensional space (such as range search and nearest neighbor search). Preferably, the processing procedure of the K-Dimensional tree matching algorithm is as follows: each point in the standard real-time point cloud map is traversed in turn, and a matching point corresponding to the currently traversed point is determined from the reference point cloud map, it is judged whether the distance between the currently traversed point and the matching point is greater than the preset distance threshold, if yes, the currently traversed point is added to the difference point cloud map, and the next point in the standard real-time point cloud map is traversed; if not, the next point in the standard real-time point cloud map is traversed.

[0106] Further, in the step of implementing step S104, the points cloud representing the upper end of the hook, the points cloud representing the hook, the points cloud representing the hoisting rope and the points cloud representing the hoisted object are clustered from the difference point cloud map by using a preset clustering algorithm, specifically including:

[0107] Step B1: obtaining a clustering model trained in advance based on the Euclidean clustering algorithm;

[0108] Step B2: inputting the three-dimensional coordinate position of each point in the difference point cloud map in the tower crane coordinate system into the clustering model to obtain the points cloud representing the upper end of the hook, the points cloud representing the hook, the points cloud representing the hoisting rope and the points cloud representing the hoisted object.

[0109] In this embodiment, the clustering model is obtained by training based on the Euclidean clustering algorithm using preset sample data, and the points cloud representing the steel wire rope, the points cloud representing the hook, the points cloud representing the hoisting rope and the points cloud representing the hoisted object can be more accurately obtained through the clustering model.

[0110] Further, in the step of implementing step S105, the contour information of the hoisted object is calculated according to the position information of each point in the points cloud representing the hoisted object, specifically including:

[0111] Step C1: Obtain the three-dimensional coordinate position of each point in the point cloud representing the suspended object in the tower crane coordinate system, and take the difference between the maximum Z-axis coordinate and the minimum Z-axis coordinate as the height of the suspended object;

[0112] Step C2: Obtain a two-dimensional image based on the X-axis and Y-axis coordinates of each point in the point cloud representing the suspended object;

[0113] The two-dimensional graphic is a binary image, where the value of a valid point is 255 and the value of an invalid point is 0.

[0114] Step C3: Divide the two-dimensional image into N×N patches according to the preset value N, and calculate the principal direction angle of each patch;

[0115] Specifically, for the target image patch, the main direction angle of the target image patch is calculated based on the pixel values ​​of the neighboring image patches surrounding the target image patch;

[0116] Step C4: Calculate the principal direction angle of the two-dimensional image based on the principal direction angles of all the tiles;

[0117] Furthermore, step C4 specifically includes:

[0118] Step C41: Based on the main direction angle of each tile, classify each tile into a preset M angle range; wherein, the 360 ​​degrees are evenly divided into M angle ranges according to the preset value M;

[0119] Step C42: Count the number of tiles in each angle range in turn, and take the angle range with the most tiles as the main direction angle of the two-dimensional image;

[0120] Step C5: Rotate the two-dimensional image according to the main direction angle of the two-dimensional image to obtain a standard two-dimensional image;

[0121] like Figure 6 The image shown is a two-dimensional image before rotation, as follows: Figure 7 The image shown is a rotated standard two-dimensional image;

[0122] Step C6: Calculate the length and width of the suspended object based on the points located in the upper left corner and the lower right corner of the standard two-dimensional image;

[0123] Furthermore, step C6 specifically includes:

[0124] Step C61: Obtain the coordinates (left, top) of the point located at the top left corner in the standard two-dimensional image;

[0125] Step C62: Obtain the coordinates of the point located at the lower right corner in the standard two-dimensional image (right, bottom) ;

[0126] Step C63: Calculate the length of the hoisted object length according to the following formula:

[0127] length = right-left;

[0128] Step C64: Calculate the width of the hoisted object width according to the following formula:

[0129] width = bottom-bottom.

[0130] In this embodiment, by obtaining the static point cloud map of the tower crane and the real-time point cloud map of the laser radar, and performing point cloud registration, the difference between the current tower crane trolley lower environment range and the static point cloud map is automatically calculated in real time, so as to output the difference point cloud map of the ROI region (steel wire rope, hook, sling, hoisted object) in the tower crane coordinate system; In this embodiment, through real-time acquisition and algorithm calculation of point cloud data, the contour information of the hoisted object can be automatically calculated, and manual input is avoided.

[0131] Embodiment two

[0132] The embodiment of the application provides a device for identifying the contour of a hoisted object, as shown in the figure, which specifically comprises the following components: Figure 8

[0133] The first acquisition module 801 is configured to acquire a static point cloud map representing the ground information in the working area of the tower crane when the tower crane is not in use.

[0134] The second acquisition module 802 is configured to acquire a real-time point cloud map representing the area below the trolley of the tower crane when the tower crane is in use.

[0135] The point cloud comparison module 803 is configured to compare the static point cloud map and the real-time point cloud map, and form a difference point cloud map from the points only existing in the real-time point cloud map but not existing in the static point cloud map.

[0136] The point cloud clustering module 804 is configured to cluster the point cloud representing the upper end steel wire rope of the hook, the point cloud representing the hook, the point cloud representing the sling, and the point cloud representing the hoisted object from the difference point cloud map by using a preset clustering algorithm.

[0137] The contour identification module 805 is configured to calculate the contour information of the hoisted object according to the position information of each point in the point cloud representing the hoisted object.

[0138] ​Specifically, the static point cloud map, the real-time point cloud map and the difference point cloud map all adopt a preset tower crane coordinate system; wherein the tower crane coordinate system is established with the intersection of the tower crane and the ground as the origin, the positive east direction as the X axis, the positive north direction as the Y axis, and the upward direction perpendicular to the ground as the Z axis.

[0139] Further, the point cloud comparison module 803 specifically comprises:

[0140] an acquisition unit, configured to acquire a current coordinate position of the tower crane trolley in the tower crane coordinate system;

[0141] a cutting unit, configured to cut a reference point cloud map from the static point cloud map with the current coordinate position as a positioning point; wherein the ground area represented by the reference point cloud map is the same as the ground area represented by the real-time point cloud map;

[0142] a comparison unit, configured to compare the reference point cloud map and the real-time point cloud map, and form the difference point cloud map from the points existing only in the real-time point cloud map but not in the reference point cloud map.

[0143] Further, the acquisition unit is specifically configured to:

[0144] acquire an angle γ between the initial position of the tower crane jib and the positive east direction, and acquire a rotation angle θ of the tower crane jib when the tower crane is in use;

[0145] acquire a distance r of the tower crane trolley from the origin of the tower crane coordinate system in a top view, and acquire a ground clearance h of the tower crane jib;

[0146] calculate the current position (x, y, z) of the tower crane trolley in the tower crane coordinate system according to the following formula:

[0147] x = r * cos (γ + θ);

[0148] y = r * sin (γ + θ);

[0149] z = h.

[0150] Further, the cutting unit is specifically configured to:

[0151] set the X axis coordinate and the Y axis coordinate in the current coordinate position as reference points, and set a reference radius according to the size of the ground area corresponding to the real-time point cloud map;

[0152] form a first set of the X axis coordinate and the Y axis coordinate of each point in the static point cloud map, and create an empty second set;

[0153] sequentially traverse each point in the first set, and determine whether the distance between the current traversed point and the reference point is less than the reference radius;

[0154] If yes, the current traversed point is added to the second set, and if no, the next point in the first set is traversed;

[0155] When all points in the first set are traversed, the reference point cloud map is formed according to all points in the second set.

[0156] Further, the comparison unit is specifically configured to:

[0157] input the reference point cloud map and the real-time point cloud map into an Iterative Closest Point (ICP) algorithm to obtain a rotation parameter R* and a translation parameter t* for aligning and adjusting the real-time point cloud map;

[0158] align and adjust the real-time point cloud map according to the following formula to obtain a standard real-time point cloud map:

[0159] Standard real-time point cloud map = Real-time point cloud map × R * +t * ;

[0160] input the standard real-time point cloud map, the reference point cloud map and a preset distance threshold into a K-Dimensional tree matching algorithm to determine the difference point cloud map.

[0161] Further, the point cloud clustering module 804 is specifically configured to:

[0162] obtain a clustering model trained in advance based on a Euclidean clustering algorithm;

[0163] input the three-dimensional coordinate position of each point in the difference point cloud map in the tower crane coordinate system into the clustering model to obtain point cloud representing the upper end steel wire rope of the hook, point cloud representing the hook, point cloud representing the hoisting rope, and point cloud representing the hoisted object.

[0164] Further, the contour recognition module 805 is specifically configured to:

[0165] obtain the three-dimensional coordinate position of each point in the point cloud representing the hoisted object in the tower crane coordinate system, and take the difference between the maximum Z-axis coordinate and the minimum Z-axis coordinate as the height of the hoisted object;

[0166] obtain a two-dimensional image according to the X-axis coordinate and the Y-axis coordinate of each point in the point cloud representing the hoisted object;

[0167] divide the two-dimensional image into N×N blocks according to a preset value N, and calculate the main direction angle of each block.

[0168] calculating a main direction angle of the two-dimensional image according to the main direction angles of all the tiles;

[0169] rotating the two-dimensional image according to the main direction angle of the two-dimensional image to obtain a standard two-dimensional image;

[0170] calculating the length and width of the hoist according to a point located at the upper left corner and a point located at the lower right corner of the standard two-dimensional image.

[0171] Embodiment Three

[0172] The embodiment also provides a computer device, such as a smartphone, a tablet computer, a notebook computer, a desktop computer, a rack server, a blade server, a tower server, or a cabinet server (including a single server or a server cluster composed of multiple servers), etc. that can execute programs. As shown in the figure, the computer device 90 of the embodiment at least includes but is not limited to a memory 901 and a processor 902 that can be connected to each other in communication through a system bus. It should be noted that, Figure 9 only the computer device 90 with components 901-902 is shown, but it should be understood that all the components shown are not required to be implemented, and more or fewer components can be alternatively implemented. Figure 9 only the computer device 90 with components 901-902 is shown, but it should be understood that all the components shown are not required to be implemented, and more or fewer components can be alternatively implemented.

[0173] In the embodiment, the memory 901 (i.e., a readable storage medium) includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 901 can be an internal storage unit of the computer device 90, such as a hard disk or a memory of the computer device 90. In other embodiments, the memory 901 can also be an external storage device of the computer device 90, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 90. Of course, the memory 901 can include both the internal storage unit and the external storage device of the computer device 90. In the embodiment, the memory 901 is generally used to store an operating system and various application software installed on the computer device 90. In addition, the memory 901 can also be used to temporarily store various data that have been output or will be output.

[0174] The processor 902 may, in some embodiments, be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 902 is generally used to control the overall operation of the computer device 90.

[0175] Specifically, in the present embodiment, the processor 902 is configured to execute a program of a method of identifying a profile of a hoisted object stored in the memory 901, the program of the method of identifying a profile of a hoisted object, when executed, implements the following steps:

[0176] acquiring a static point cloud map representing ground information within a working area of the tower crane when the tower crane is not in use;

[0177] acquiring a real-time point cloud map representing an area below a trolley of the tower crane when the tower crane is in use;

[0178] comparing the static point cloud map and the real-time point cloud map, and forming a difference point cloud map from points only existing in the real-time point cloud map but not existing in the static point cloud map;

[0179] clustering, by using a preset clustering algorithm, points representing a wire rope at an upper end of a hook, points representing the hook, points representing a hoist rope, and points representing the hoisted object from the difference point cloud map;

[0180] calculating, according to position information of each point in the points representing the hoisted object, profile information of the hoisted object.

[0181] The specific embodiment process of the above method steps can be referred to Embodiment One, which will not be repeated here.

[0182] Embodiment Four

[0183] The present embodiment also provides a computer readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, a server, an App application store, etc., having a computer program stored thereon, the computer program, when executed by a processor, implements the following method steps:

[0184] acquiring a static point cloud map representing ground information within a working area of the tower crane when the tower crane is not in use;

[0185] acquiring a real-time point cloud map representing an area below a trolley of the tower crane when the tower crane is in use;

[0186] Comparing the static point cloud map and the real-time point cloud map, points only existing in the real-time point cloud map but not existing in the static point cloud map form a difference point cloud map;

[0187] Clustering points representing the upper end wire rope of the hook, points representing the hook, points representing the hoisting rope, and points representing the hoisted object from the difference point cloud map by using a preset clustering algorithm;

[0188] According to position information of each point in the points representing the hoisted object, contour information of the hoisted object is calculated.

[0189] The specific embodiment process of the above method steps can be referred to Embodiment 1, which is not repeated here.

[0190] It should be noted that in this article, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without more limitations, the element defined by the statement "including a" does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0191] The above-mentioned embodiment number of the present application is only for description, not representing the pros and cons of the embodiments.

[0192] Through the above description of the embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment method can be realized by means of software and necessary general hardware platform, of course, it can also be realized by hardware, but in many cases, the former is a better embodiment.

[0193] The above is only the preferred embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields, which is made by using the content of the present application specification and drawings, is also included in the patent protection scope of the present application.

Claims

1. A method of identifying a profile of a load, characterized by, The method comprises: acquiring a static point cloud map representing ground information in a working area of a tower crane when the tower crane is not in use; acquiring a real-time point cloud map representing an area below a trolley of the tower crane when the tower crane is in use; comparing the static point cloud map and the real-time point cloud map, and forming a difference point cloud map from points existing only in the real-time point cloud map but not in the static point cloud map; clustering, from the difference point cloud map, point clouds representing a steel wire rope at an upper end of a hook, point clouds representing the hook, point clouds representing a sling, and point clouds representing a hoisted object, by using a preset clustering algorithm; calculating, according to position information of each point in the point clouds representing the hoisted object, contour information of the hoisted object; wherein the calculating, according to position information of each point in the point clouds representing the hoisted object, contour information of the hoisted object comprises: acquiring three-dimensional coordinate positions of each point in the point clouds representing the hoisted object in a preset tower crane coordinate system, and taking a difference between a maximum Z-axis coordinate and a minimum Z-axis coordinate as a height of the hoisted object; wherein the static point cloud map, the real-time point cloud map, and the difference point cloud map all adopt the tower crane coordinate system; obtaining a two-dimensional image according to X-axis coordinates and Y-axis coordinates of each point in the point clouds representing the hoisted object; dividing the two-dimensional image into N×N blocks according to a preset value N, and calculating a main direction angle of each block; classifying each block into a preset M angle range according to the main direction angle of each block; wherein 360 degrees are evenly divided into M angle ranges according to a preset value M; sequentially counting the number of blocks in each angle range, and taking an angle range with the largest number of blocks as a main direction angle of the two-dimensional image; rotating the two-dimensional image according to the main direction angle of the two-dimensional image to obtain a standard two-dimensional image; calculating a length and a width of the hoisted object according to a point located at a top left corner and a point located at a bottom right corner in the standard two-dimensional image.

2. The method of identifying a hoist profile according to claim 1, wherein, The tower crane coordinate system is established with an intersection of the tower crane and the ground as an origin, a positive east direction as an X-axis, a positive north direction as a Y-axis, and a direction perpendicular to the ground and upward as a Z-axis.

3. The method of identifying a hoist profile according to claim 2, wherein, The comparing the static point cloud map and the real-time point cloud map, and forming a difference point cloud map from points existing only in the real-time point cloud map but not in the static point cloud map comprises: acquiring a current coordinate position of the trolley in the tower crane coordinate system; taking the current coordinate position as a positioning point to intercept a reference point cloud map from the static point cloud map; wherein a ground area represented by the reference point cloud map is the same as a ground area represented by the real-time point cloud map; comparing the reference point cloud map and the real-time point cloud map, and forming the difference point cloud map from points existing only in the real-time point cloud map but not in the reference point cloud map.

4. The method of identifying a hoist profile according to claim 3, wherein, The taking the current coordinate position as a positioning point to intercept a reference point cloud map from the static point cloud map comprises: setting X-axis coordinates and Y-axis coordinates in the current coordinate position as reference points, and setting a reference radius according to a size of a ground area corresponding to the real-time point cloud map. form a first set with X-axis coordinates and Y-axis coordinates of each point in the static point cloud map, and create an empty second set; sequentially traverse each point in the first set, and determine whether the distance between the current traversed point and the reference point is less than the reference radius; if yes, add the current traversed point to the second set, and if no, traverse the next point in the first set; when all points in the first set are traversed, form the reference point cloud map according to all points in the second set.

5. The method of identifying a hoist profile according to claim 3, wherein, The comparison of the reference point cloud map and the real-time point cloud map forms the difference point cloud map only from points existing in the real-time point cloud map but not in the reference point cloud map, which includes: inputting the reference point cloud map and the real-time point cloud map into an iterative closest point (ICP) algorithm to obtain a rotation parameter R* and a translation parameter t* for aligning and adjusting the real-time point cloud map; aligning and adjusting the real-time point cloud map according to the following formula to obtain a standard real-time point cloud map: Standard real-time point cloud map = Real-time point cloud map x + ; inputting the standard real-time point cloud map, the reference point cloud map and a preset distance threshold into a K-Dimensional tree matching algorithm to determine the difference point cloud map.

6. The method of identifying a hoist profile according to claim 2, wherein, The clustering of the difference point cloud map by using a preset clustering algorithm to obtain point clouds representing the upper end of the steel wire rope of the hook, point clouds representing the hook, point clouds representing the rope and point clouds representing the hoisted object, which includes: obtaining a clustering model trained in advance based on a Euclidean clustering algorithm; inputting three-dimensional coordinate positions of each point in the difference point cloud map in the tower crane coordinate system into the clustering model to obtain point clouds representing the upper end of the steel wire rope of the hook, point clouds representing the hook, point clouds representing the rope and point clouds representing the hoisted object.

7. A device for identifying a profile of a suspended object, characterized in that The device includes: a first obtaining module configured to obtain a static point cloud map representing ground information in a working area of a tower crane when the tower crane is not in use; a second obtaining module configured to obtain a real-time point cloud map representing an area below a trolley of the tower crane when the tower crane is in use; a point cloud comparison module configured to compare the static point cloud map and the real-time point cloud map, and form a difference point cloud map from points existing only in the real-time point cloud map but not in the static point cloud map; a point cloud clustering module configured to cluster point clouds representing the upper end of the steel wire rope of the hook, point clouds representing the hook, point clouds representing the rope and point clouds representing the hoisted object from the difference point cloud map by using a preset clustering algorithm; a contour identification module configured to calculate contour information of the hoisted object according to position information of each point in the point clouds representing the hoisted object; The contour identification module is configured to: obtain three-dimensional coordinate positions of each point in the point clouds representing the hoisted object in a preset tower crane coordinate system, and take a difference between a maximum Z-axis coordinate and a minimum Z-axis coordinate as a height of the hoisted object; wherein the static point cloud map, the real-time point cloud map and the difference point cloud map all adopt the tower crane coordinate system; obtain a two-dimensional image according to X-axis coordinates and Y-axis coordinates of each point in the point clouds representing the hoisted object; and According to a preset value N, the two-dimensional image is divided into N×N blocks, and a main direction angle of each block is calculated; According to the main direction angle of each block, each block is classified into a preset M angle ranges; wherein 360 degrees are equally divided into M angle ranges according to a preset value M; The number of blocks in each angle range is counted in sequence, and the angle range with the largest number of blocks is taken as the main direction angle of the two-dimensional image; The two-dimensional image is rotated according to the main direction angle of the two-dimensional image to obtain a standard two-dimensional image; The length and width of the hanging object are calculated according to a point located at the upper left corner and a point located at the lower right corner in the standard two-dimensional image.

8. A computer device comprising: Memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of any one of claims 1 to 6 when executing the computer program.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 6.

Images