CN115546300A - Method and device for identifying pose of tray placed tightly, computer equipment and medium - Google Patents
Method and device for identifying pose of tray placed tightly, computer equipment and medium Download PDFInfo
- Publication number
- CN115546300A CN115546300A CN202211239101.5A CN202211239101A CN115546300A CN 115546300 A CN115546300 A CN 115546300A CN 202211239101 A CN202211239101 A CN 202211239101A CN 115546300 A CN115546300 A CN 115546300A
- Authority
- CN
- China
- Prior art keywords
- tray
- point cloud
- cloud data
- target
- original
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
- G06V10/763—Non-hierarchical techniques, e.g. based on statistics of modelling distributions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The application relates to a method and a device for identifying the pose of a tray which is placed closely, computer equipment and a medium. The method comprises the following steps: acquiring original 3D point cloud data of a target goods taking area; the target goods taking area is used for placing trays to be taken, a plurality of trays to be taken are placed closely, and the trays to be taken comprise a plurality of tray foot piers; selecting a plurality of target points from the original 3D point cloud data, and projecting the target points to obtain an original tray image; extracting target point cloud data corresponding to the tray to be taken from original 3D point cloud data according to the original tray image; and calculating the pose information of the tray to be taken according to the target point cloud data. By adopting the method, the accurate positioning and identification of the tray pose can be realized.
Description
Technical Field
The application relates to the technical field of forklifts, in particular to a method and a device for identifying pose of a tray which is placed closely, computer equipment and a medium.
Background
Fork lift trucks are various wheeled transportation vehicles for loading and unloading, stacking, short-distance transportation and heavy load transportation of pallet loads, which are important transportation tools for modern logistics transportation. The tray can be widely applied to ports, stations, airports, goods yards, factory workshops, warehouses, circulation centers, distribution centers and the like, can enter cabins, carriages and containers to carry out loading, unloading and carrying operations of tray goods, and is necessary equipment in tray transportation and container transportation. The key of the forklift for carrying is how to position the pose of the pallet.
At present, the tray is mainly identified and positioned by a 2d laser range finder, however, characteristic information captured from 2d depth information only has distance and length, only the pier hole length on one plane can be considered, the whole tray cannot be extended, and the position and posture of the tray are not accurate. Therefore, how to realize the accurate positioning of the position and posture of the pallet to be taken by the forklift becomes a technical problem which needs to be solved urgently by technical personnel in the field.
Disclosure of Invention
Therefore, in order to solve the technical problems, a compact tray pose identification method and apparatus, a computer device, and a medium are provided, which can accurately position the pose of a tray to be picked.
In a first aspect, the application provides a method for identifying poses of trays placed tightly. The method comprises the following steps:
acquiring original 3D point cloud data of a target goods taking area; the target goods taking area is provided with a plurality of trays to be taken, the plurality of trays to be taken are tightly placed, and the trays to be taken comprise a plurality of tray foot piers;
selecting a plurality of target points from the original 3D point cloud data, and projecting the target points to obtain an original tray image;
extracting target point cloud data corresponding to the tray to be taken from original 3D point cloud data according to the original tray image;
and calculating the pose information of the tray to be taken according to the target point cloud data.
In one embodiment, the selecting a plurality of target points from the raw 3D point cloud data comprises:
filtering the original 3D point cloud data according to a preset target position to obtain first point cloud data;
and carrying out voxel filtering on the first point cloud data to obtain a plurality of target points.
In one embodiment, the extracting, from original 3D point cloud data according to the original tray image, target point cloud data corresponding to the tray to be taken includes:
carrying out image recognition processing on the original tray image to obtain a tray recognition area; the tray identification area is used for representing an area of the tray to be taken in the original tray image;
filtering the original 3D point cloud data according to the tray identification area to obtain filtered point cloud data;
and fitting the filtered point cloud data to obtain target point cloud data corresponding to the tray to be taken.
In one embodiment, the performing image recognition processing on the original tray image to obtain a tray recognition area includes:
performing target recognition on the original tray image according to a preset target recognition algorithm to obtain a plurality of target recognition areas; wherein each target identification area comprises image position information of a tray in an original tray image;
calculating to obtain central position information corresponding to each target identification area according to the image position information;
screening each target identification area according to a preset target position and the central position information to obtain a tray identification area; wherein the tray identification area corresponds to the tray to be taken.
In one embodiment, the performing target recognition on the original tray image according to a preset target recognition algorithm to obtain a plurality of target recognition areas includes:
carrying out grid division on the original tray image to obtain a plurality of grid units;
carrying out target detection on each grid unit to obtain a plurality of target detection areas corresponding to each grid unit;
calculating to obtain the boundary overlapping degree corresponding to each target detection area according to the plurality of target detection areas corresponding to each grid unit and a preset target mark area;
and screening the target detection area according to the overlapping degree threshold and the boundary overlapping degree to obtain the target identification area.
In one embodiment, the calculating the pose information of the tray to be taken according to the target point cloud data includes:
dividing the target point cloud data into a plurality of second point cloud data; wherein each of the second point cloud data comprises a plurality of point clouds;
ordering the second point cloud data to obtain ordered second point cloud data, and performing Euclidean clustering on each ordered second point cloud data according to a preset clustering radius to obtain a clustering result corresponding to each second point cloud data; the clustering result is used for representing whether the second point cloud data comprises all the tray foot piers;
if the clustering result corresponding to the second point cloud data represents that the second point cloud data comprises all tray foot piers, acquiring coordinates corresponding to each point cloud in the second point cloud data to obtain foot pier coordinate data;
and calculating the pose information of the tray to be taken according to the foot pier coordinate data.
In one embodiment, the performing the euclidean clustering on each second point cloud data according to a preset clustering radius to obtain a clustering result corresponding to each second point cloud data includes:
performing Euclidean clustering on each second point cloud data according to the clustering radius to obtain the clustering number, the clustering width and the clustering distance of clustering bodies corresponding to each second point cloud data;
and if the clustering number of the clustering bodies is larger than or equal to the number of the tray piers corresponding to the tray pier, the difference value between the clustering width and the pier width threshold value is smaller than a width difference threshold value, and the clustering distance between the nearest clustering bodies is smaller than a clamping plate error threshold value, the clustering result represents that the second point cloud data comprises all the tray piers.
In a second aspect, the application further provides a pallet pose recognition device which is placed closely. The device comprises:
the point cloud acquisition module is used for acquiring original 3D point cloud data of a target goods taking area; the target goods taking area is used for placing trays to be taken, a plurality of trays to be taken are placed closely, and the trays to be taken comprise a plurality of tray foot piers;
the image projection module is used for selecting a plurality of target points from the original 3D point cloud data and projecting the target points to obtain an original tray image;
the point cloud extraction module is used for extracting target point cloud data corresponding to the tray to be taken from original 3D point cloud data according to the original tray image;
and the pose calculation module is used for calculating pose information of the tray to be taken according to the target point cloud data.
In a third aspect, the present application also provides a computer device. The computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the method for identifying the pose of the tightly placed tray in the embodiment of the first aspect when executing the computer program.
In a fourth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium, on which a computer program is stored, which, when executed by a processor, implements the method for identifying a pose of a tightly set tray as described in the embodiments of the first aspect.
According to the method and the device for identifying the pose of the tray which is placed tightly, the computer equipment and the medium, original 3D point cloud data of a target goods taking area are obtained, wherein the tray to be taken is placed in the target goods taking area, a plurality of trays to be taken are placed tightly, the tray to be taken comprises a plurality of tray foot piers, a plurality of target points are selected from the original 3D point cloud data and projected to obtain an original tray image, target point cloud data corresponding to the tray to be taken are extracted from the original 3D point cloud data according to the original tray image, and finally pose information of the tray to be taken is obtained through calculation according to the target point cloud data. According to the technical scheme, the position and posture identification of the tray to be taken is achieved by combining the original 3D point cloud data and the original tray image, and the accuracy of the position and posture identification of the tray to be taken is improved.
Drawings
FIG. 1 is a schematic flow diagram of a method for identifying pose of a tightly packed pallet in some embodiments;
FIG. 2 is a schematic diagram of the relationship of a pixel coordinate system, a laser coordinate system, and a voxel coordinate system in some embodiments;
FIG. 3 is a schematic flow chart of an embodiment of the method of step 106 in FIG. 1;
FIG. 4 is a schematic flow chart of an embodiment of the method of step 302 in FIG. 3;
FIG. 5 is a flowchart illustrating a specific method of step 402 in FIG. 4;
FIG. 6 is a schematic flow chart of an embodiment of the method of step 108 in FIG. 1;
FIG. 7 is a flowchart illustrating an embodiment of the method of step 604 in FIG. 6;
FIG. 8 is a flow chart illustrating a method for identifying poses of tightly packed pallets in further embodiments;
FIG. 9 is a block diagram of a compact pallet pose identification apparatus in some embodiments;
FIG. 10 is a diagram showing an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
In one embodiment, as shown in fig. 1, a method for identifying the pose of a tray placed tightly is provided, and this embodiment is exemplified by applying the method to a server, and it is understood that the method may also be applied to a handling apparatus, and may also be applied to a system including a handling apparatus and a server, and is implemented by interaction between the handling apparatus and the server. Wherein, the handling device refers to a transportation device for handling goods, and the handling device may be, but is not limited to, at least one of an automatic Guided Vehicle (AGV cart) and a forklift, wherein the forklift may be, but is not limited to, an unmanned forklift. The method for identifying the pose of the tray which is placed tightly comprises the following steps:
102, acquiring original 3D point cloud data of a target goods taking area; the target goods taking area is provided with a tray to be taken, a plurality of trays to be taken are tightly placed, and the tray to be taken comprises a plurality of tray foot piers.
A pallet is a cargo vehicle for transporting goods in groups.
The tray to be taken refers to a tray which is specified by the object goods taking area and needs to be carried by the carrying equipment to take goods.
Point cloud data refers to a set of vectors in a three-dimensional coordinate system, the vectors are usually represented in the form of X, Y, Z three-dimensional coordinates, and the point cloud data may also contain color information and reflection intensity information.
The target goods taking area in the step 102 refers to an area where the carrying equipment forks the tray, a plurality of trays to be taken are placed in the target goods taking area, and goods to be taken can be placed on the trays to be taken or not. Each tray to be taken comprises a tray foot pier.
The original 3D point cloud data may be obtained by a 3D laser, or may be obtained by other means, which is not specifically limited in this application. As obtained by 3D laser acquisition, the 3D laser may be mounted on a handling device. For example, a 3D laser may be installed at a mid-point between the fork arms of the forklift to obtain raw 3D point cloud data of the target pickup area through the 3D laser acquisition. Of course, the 3D laser may be installed at other positions of the forklift or at other positions of the target pickup area, and the application is not particularly limited thereto.
In some embodiments, the 3D laser is installed at the middle point between the fork arms of the forklift, so that the laser equipment can be guaranteed not to be shielded by goods during laser scanning, and the laser scanning accuracy is improved. The original 3D point cloud data obtained by the 3D laser is located in a laser coordinate system, and the origin of the laser coordinate system may be set at the installation position of the 3D laser or at other positions.
And 104, selecting a plurality of target points from the original 3D point cloud data, and projecting the target points to obtain an original tray image.
In step 104, because the original 3D point cloud data is in the laser coordinate system, one of the coordinate systems may be selected to screen the original 3D point cloud data, so as to select and obtain a plurality of target points, and then the surface is selected as a projection surface to perform projection processing on the plurality of target points, so as to obtain an original tray image.
In some embodiments, the projecting the plurality of target points to obtain the original tray image may include:
and projecting the target points to obtain a preliminary projection image.
And carrying out gray level processing on the preliminary projection image to obtain an original tray image.
As shown in fig. 2, fig. 2 is a schematic diagram of a relationship among a pixel coordinate system, a laser coordinate system, and a voxel coordinate system according to an embodiment of the present application. In the present embodimentProjecting a plurality of target points to obtain a preliminary projection image, wherein the preliminary projection image is an RGB image, establishing a pixel coordinate system for the preliminary projection image, the origin of the pixel coordinate system is consistent with the origin of a voxel coordinate system, and the origin of the pixel coordinate system is (x) in the voxel coordinate system min ,y min ) The point is (x _ min, y _ min) in the figure. For each pixel point in the pixel coordinate system, if a target point exists in the pixel point, the corresponding pixel gray value is 255, and if the target point does not exist in the pixel point, the corresponding pixel gray value is 0. Through the arrangement, the graying of the primary projection image is realized, and the original tray image is obtained. The arrangement enables the original tray image to only comprise information related to the tray to be taken, thereby facilitating the subsequent realization of the position and posture identification of the tray to be taken and improving the accuracy of the position and posture identification of the tray to be taken.
And 106, extracting target point cloud data corresponding to the tray to be taken from the original 3D point cloud data according to the original tray image.
In step 106, determining an approximate region where the tray to be taken is located through the original tray image, and after determining the approximate region where the tray to be taken is located, extracting point cloud data corresponding to the approximate region from the original 3D point cloud data, so as to obtain target point cloud data corresponding to the tray to be taken.
And 108, calculating to obtain the pose information of the tray to be taken according to the target point cloud data.
In step 108, after the target point cloud data corresponding to the tray to be taken is obtained through the above steps, the pose information of the tray to be taken can be obtained through calculation according to the target point cloud data, the pose information can be represented by x, y, z and yaw of the tray to be taken, and yaw represents yaw information.
And after the position and the posture information of the tray are obtained, the carrying equipment starts a goods taking task.
The tray pose identification method is always executed in the process of taking the goods from the tray by the carrying equipment.
According to the method for identifying the pose of the tightly placed tray, original 3D point cloud data of a target goods taking area are obtained, wherein a tray to be taken is placed in the target goods taking area, a plurality of trays to be taken are placed tightly, the tray to be taken comprises a plurality of tray foot piers, a plurality of target points are selected from the original 3D point cloud data and projected to obtain an original tray image, target point cloud data corresponding to the tray to be taken is extracted from the original 3D point cloud data according to the original tray image, and finally pose information of the tray to be taken is obtained through calculation according to the target point cloud data. Through the arrangement, the position and pose identification of the tray to be taken is realized by combining the original 3D point cloud data and the original tray image, and the position and pose identification accuracy of the tray to be taken is improved.
In some embodiments, step 104 includes the steps of: filtering the original 3D point cloud data according to a preset target position to obtain first point cloud data; and carrying out voxel filtering on the first point cloud data to obtain a plurality of target points.
The target position refers to an ideal coordinate pose of the tray to be taken relative to a coordinate system where the carrying device is located, and specifically includes but is not limited to at least one of a longitudinal width of the tray to be taken relative to the coordinate system where the carrying device is located in an X-axis direction, a transverse length of the tray to be taken relative to the coordinate system where the carrying device is located in a Y-axis direction, a vertical height of the tray to be taken relative to the coordinate system where the carrying device is located in a Z-axis direction, and an included angle of a tray plane of the tray to be taken relative to the coordinate system where the carrying device is located in the Y-axis direction.
In some embodiments, the target bit may be in x 0 ,y 0 ,z 0 ,th 0 Is represented by, wherein x 0 Indicating the longitudinal width, y, of the pallet to be taken in the direction of the X-axis relative to the coordinate system of the handling device 0 Representing the transverse length of the pallet to be picked in the Y-axis direction with respect to the coordinate system of the handling apparatus, z 0 Indicating the vertical height, th, of the pallet to be taken in the Z-axis direction relative to the coordinate system of the handling device 0 And the included angle of the plane of the tray to be taken and the coordinate system of the carrying equipment in the Y-axis direction is shown.
And when the carrying equipment runs to the target goods taking area, the carrying equipment issues the target position of the tray to be taken to the perception algorithm module. And the perception algorithm module is a module used for filtering point cloud data on the server.
In some embodiments, the process of filtering the original 3D point cloud data to obtain the first point cloud data by the perceptual algorithm module according to the target bits is as follows: the perception algorithm module obtains laser external parameters calibrated in advance and tray size parameters of a tray to be taken configured on the perception algorithm module in advance, and point cloud data matched with the size of the tray to be taken are extracted from the original 3D point cloud data according to the tray size parameters and serve as first point cloud data.
The laser external reference refers to a coordinate system conversion relation between the laser device and other coordinate systems, for example, a coordinate system where the carrying device is located. Tray size parameters including, but not limited to, at least one of tray length, tray width, and tray height.
In some embodiments, the voxel filtering of the XOY plane in the laser coordinate system is performed on the first point cloud data, specifically:
and in a cuboid voxel grid with a preset size, replacing all points in each cuboid voxel grid with the point with the maximum y value under a laser coordinate system to obtain a plurality of target points. That is, for each cubic voxel grid, the point with the maximum y value in the voxel grid is selected as the corresponding target point. Wherein each cube may be set to 1cm by 1cm. Of course, the setting of other sizes is also possible, and the setting can be performed according to actual situations, and the application is not particularly limited.
Referring to fig. 3, in some embodiments, step 106 includes, but is not limited to, the following steps:
And 304, filtering the original 3D point cloud data according to the tray identification area to obtain filtered point cloud data.
In step 304 of some embodiments, after the tray identification area corresponding to the tray to be taken is obtained according to the foregoing steps, the original 3D point cloud data is filtered through the tray identification area to obtain target point cloud data, and the target point cloud data is used for representing the point cloud data corresponding to the tray to be taken.
Specifically, the original 3D point cloud data may be subjected to straight-through filtering processing to obtain filtered point cloud data.
Step 306: and fitting the filtered point cloud data to obtain target point cloud data corresponding to the tray to be taken.
In some embodiments, the filtered point cloud data is subjected to a plane fitting process, and points that meet a set distance are extracted to obtain target point cloud data. The set distance is preset, and may be modified according to actual conditions, and the present application is not limited specifically.
Specifically, a Random Sample Consensus (Random Sample Consensus) algorithm may be adopted to randomly take points for fitting, a set distance may be 2cm or 3cm, and a plane fitting is performed on the filtered point cloud data by using the Random Sample Consensus algorithm, so as to extract points that meet the set distance to obtain target point cloud data.
In some embodiments, step 304 includes the steps of:
maximum and minimum values of x of the tray identification area are acquired, and maximum and minimum values of y of the tray identification area are acquired.
And performing through filtering processing on the original 3D point cloud data according to the maximum value and the minimum value of x of the tray identification area and the maximum value and the minimum value of y of the tray identification area to obtain filtered point cloud data.
Specifically, a central coordinate point of a tray identification area is obtained, wherein the central coordinate point is the central coordinate point of the tray identification area under a pixel coordinate system, the central coordinate point is converted under a laser coordinate system, then, the maximum value and the minimum value of the abscissa of the tray identification area under the laser coordinate system are obtained, the maximum value and the minimum value of the ordinate of the tray identification area are obtained, and then, the original 3D point cloud data are filtered in a range limited by the maximum value and the minimum value of the abscissa of the tray identification area and the maximum value and the minimum value of the ordinate of the tray identification area, so that the target point cloud data can be obtained.
Referring to FIG. 4, in some embodiments, step 302 includes, but is not limited to, the following steps:
Specifically, in step 402 of some embodiments, the target recognition algorithm selects a YOLO algorithm, and performs target recognition on the original tray image using the YOLO algorithm to identify a plurality of target recognition areas where the tray to be picked exists. Each target recognition area comprises image position information of the tray in the original tray image, at least one tray to be taken may exist in each target recognition area, and the image position information refers to the position area of the tray in the original tray image.
In some embodiments, the target identification area is a target aim frame, and the image location information, the confidence of target matching, and the IOU value are represented by pixel coordinate points of the target aim frame in a pixel coordinate system. Wherein the confidence of the target match characterizes the likelihood of the target match. For example, the image position information is represented by pixel coordinate points at the upper left and lower right corners of the target frame.
YoLO (You Only Look one: unified, real-Time Object Detection) is an Object Detection algorithm that can immediately identify objects in an image, as well as the location and relative position of the objects.
And step 404, calculating to obtain central position information corresponding to each target identification area according to the image position information.
In step 404 of some embodiments, the image position information indicates a position area of the tray in the original tray image, and then the central position information corresponding to each target identification area may be calculated according to the image position information. For example, when the pixel coordinate points of the upper left corner and the lower right corner of the target frame are used to represent the image position informationWhen the pixel coordinate point is (x) in the upper left corner min ,y min ) The pixel coordinate point at the lower right corner is expressed by (x) max ,y max ) The coordinate point of the pixel corresponding to the central position information corresponding to the tray is shown as
In step 406 of some embodiments, the target position refers to an ideal coordinate pose of the tray to be picked with respect to the coordinate system of the handling apparatus, and specifically includes, but is not limited to, at least one of a longitudinal width of the tray to be picked with respect to the coordinate system of the handling apparatus in the X-axis direction, a transverse length of the tray to be picked with respect to the coordinate system of the handling apparatus in the Y-axis direction, a vertical height of the tray to be picked with respect to the coordinate system of the handling apparatus in the Z-axis direction, and an angle of a tray plane of the tray to be picked with respect to the coordinate system of the handling apparatus in the Y-axis direction.
In some embodiments, the target bit may be in (x) 0 ,y 0 ,z 0 ,th 0 ) Is represented by, wherein x 0 Indicating the longitudinal width, y, of the pallet to be taken in the direction of the X-axis relative to the coordinate system of the handling device 0 Representing the transverse length of the pallet to be picked in the Y-axis direction with respect to the coordinate system of the handling apparatus, z 0 Indicating the vertical height, th, of the pallet to be taken in the Z-axis direction relative to the coordinate system of the handling device 0 A pixel coordinate point for indicating the included angle of the tray plane of the tray to be taken and the coordinate system of the carrying equipment in the Y-axis direction and the central position information
And (4) showing. According to the target position and the central position information, each target identification area is screened, and the specific process of obtaining the tray identification area is as follows:
go through each timeCalculating the absolute value of the difference between the pixel abscissa values of every two target identification areas, if the absolute value of the difference is greater than a preset abscissa threshold, indicating that the trays corresponding to the target identification areas are not adjacent, and in this case, selecting the x coordinate value closest to the target position 0 The corresponding target recognition area serves as a tray recognition area. If the absolute value is less than or equal to the abscissa threshold, it indicates that the tray corresponding to the target recognition area is close, in which case, the ordinate value of the pixel of the center position information of each target recognition area is traversed, and the y-coordinate value closest to the target position is selected 0 The corresponding target recognition area serves as a tray recognition area. Through the arrangement, the tray identification area corresponding to the tray to be taken can be accurately found, so that the accuracy of the position and posture identification of the tray to be taken is improved.
Referring to fig. 5, in some embodiments, step 402 specifically includes the following steps:
In step 502 of some embodiments, the original tray image is composed of a plurality of pixels, and the original tray image is in the pixel coordinate system, so that the original tray image can be directly subjected to grid division in the pixel coordinate system to obtain a plurality of grid units.
In step 504 of some embodiments, for each grid cell divided in the foregoing step, a preset fixed number of bounding boxes are selected for object detection, so as to obtain a plurality of object detection areas. The preset fixed number may be 3, or may be other numbers, and the application is not particularly limited thereto.
In step 506 of some embodiments, for multiple target detection areas in each grid cell, only the target detection area with the largest target mark area IOU is selected to be responsible for predicting whether there is a tray to be fetched. The target marking area refers to a correct result of artificial marking, and an approximate position of the tray to be taken, namely a target position issued to the perception algorithm module.
IoU (overlap over Union) is a standard that measures the accuracy with which a corresponding object is detected in a particular dataset. IoU is a simple measurement standard, and IoU can be used to measure any task that yields a prediction range (bounding boxes) in the output. The IOU may be calculated by dividing the area of the overlapping portion of the target labeling region and the target detection region by the area of the union portion of the target labeling region and the target detection region.
After the target detection area and the target mark area are obtained, the overlapping degree of each target detection area and the target mark area in each grid unit can be calculated according to the definition of the IOU, and the boundary overlapping degree corresponding to each target detection area is obtained, so that the target detection areas can be screened subsequently, the target identification area is obtained, and the accuracy of the tray pose identification is improved.
And step 508, screening the target detection area according to the overlapping degree threshold and the boundary overlapping degree to obtain a target identification area.
In step 508 of some embodiments, when the boundary overlap is greater than the overlap threshold, it indicates that the target detection area and the target mark area are a positive match, in which case the tray to be picked can be detected in the target detection area. When the boundary overlapping degree is less than or equal to the overlapping degree threshold value, it indicates that the target detection area and the target mark area are zero matching or negative matching, in which case, the tray to be taken cannot be detected in the target detection area. Therefore, the target detection area corresponding to the boundary overlapping degree greater than the overlapping degree threshold is used as the target identification area, and the target detection area corresponding to the boundary overlapping degree less than or equal to the overlapping degree threshold is screened out and not used as the target identification area.
It should be noted that the overlap threshold is preset, in some embodiments, the overlap threshold is 0.5, and certainly, the overlap threshold may also be set to other values, and may be modified according to actual situations, and the application is not limited specifically herein.
In some embodiments, as shown in FIG. 6, step 108 includes, but is not limited to, the following steps:
In step 602 of some embodiments, the target point cloud data is segmented and divided at a preset interval based on a plane of the laser coordinate system, and point cloud coordinates after segmentation and division are obtained to obtain a plurality of second point cloud data.
In some embodiments, segmentation is performed on the target point cloud data in a manner of spacing 1cm based on an XOZ plane of the laser coordinate system, and segmented point cloud coordinates are acquired, and the acquired point cloud coordinates are stored in an array to obtain a plurality of second point cloud data. For each element in the array, the element is one second point cloud data, and each second point cloud data comprises a plurality of point clouds.
Specifically, in step 604 of some embodiments, first, each point cloud data in the plurality of second point cloud data is traversed, the point clouds in each second point cloud data are sorted from small to large according to a point cloud abscissa value to obtain sorted second point cloud data, and then, according to a preset clustering radius, each second point cloud data is subjected to euclidean clustering to obtain a clustering result corresponding to each second point cloud data, where the clustering result is used to represent whether the second point cloud data includes all pallet footers of a pallet to be taken.
The objects are continuous, so that the sorted points are continuous without fail, and only the continuous points form one object, so that the sorted second point cloud data is obtained by sorting the point clouds in each second point cloud data from small to large according to the abscissa value of the point cloud, and whether the point clouds in the second point cloud data are continuous or not is judged.
Euclidean clustering is a clustering algorithm based on Euclidean distance measurement.
It should be noted that the clustering radius is preset, and may be changed according to actual situations, and the present application is not limited specifically. In some embodiments, the clustering radius is set to 5cm, and then, the second point cloud data after each sorting process is subjected to euclidean clustering according to the clustering radius of 5cm, so as to obtain a clustering result corresponding to each second point cloud data after each sorting process.
And 606, if the clustering result corresponding to the second point cloud data represents that the second point cloud data comprises all tray foot piers, obtaining coordinates corresponding to each point cloud in the second point cloud data to obtain foot pier coordinate data.
In step 606 of some embodiments, when the clustering result corresponding to the second point cloud data represents that the second point cloud data includes all tray foot piers of the tray to be taken, it indicates that the point cloud in the second point cloud data is the tray foot pier of the tray to be taken, and in this case, the coordinate corresponding to each point cloud in the second point cloud data is obtained, that is, the foot pier coordinate data corresponding to each tray foot pier in the tray to be taken can be obtained. The foot pier coordinate data is used for representing the central coordinate value of each tray foot pier in the tray to be taken.
And 608, calculating to obtain the pose information of the tray to be taken according to the foot pier coordinate data.
In step 608 of some embodiments, after the leg-pier coordinate data corresponding to all tray leg piers of the tray to be taken is obtained through calculation, the pose information of the tray to be taken can be directly obtained through calculation. The method specifically comprises the following steps:
firstly, the central coordinate value is calculated according to the foot pier coordinate data corresponding to all the tray foot piers, namely the central coordinate value of the tray to be taken can be obtained, and the central coordinate value of the tray to be taken can be expressed by (X, Y and Z).
And then calculating the difference between the adjacent tray foot piers to obtain the yaw angle of the tray to be taken. The yaw angle of the tray to be taken is calculated through a formula (1), wherein the formula (1) is as follows:
in formula (1), yaw represents the yaw angle of the tray to be taken, Z _ max and X _ max represent the coordinate value and the abscissa value in the spatial direction in the foot pier coordinate data corresponding to a certain tray foot pier respectively, and Z _ min and X _ min represent the coordinate value and the abscissa value in the spatial direction in the foot pier coordinate data corresponding to another tray foot pier respectively. Wherein the coordinate values in the spatial direction are used to express a depth in a three-dimensional cartesian coordinate system.
In some examples, in formula (1), Z _ max represents a maximum value of the coordinate value in the spatial direction in the foot pier coordinate data, X _ max represents an abscissa value of the tray foot pier corresponding to the maximum value of the coordinate value in the spatial direction in the foot pier coordinate data, and similarly, Z _ min represents a minimum value of the coordinate value in the spatial direction in the foot pier coordinate data corresponding to the tray foot pier, and X _ min represents an abscissa value of the tray foot pier corresponding to the minimum value of the coordinate value in the spatial direction in the foot pier coordinate data.
After the yaw angle and the central coordinate value of the tray to be taken are obtained through calculation, the position and attitude information of the tray to be taken can be determined, and the position and attitude information can be represented by (X, Y, Z, yaw).
In some embodiments, as shown in FIG. 7, step 604 includes, but is not limited to, the following steps:
and 702, performing Euclidean clustering on each second point cloud data according to the clustering radius to obtain the clustering number, the clustering width and the clustering distance of the clustering bodies corresponding to each second point cloud data.
In step 702 of some embodiments, the point clouds in each second point cloud data are clustered according to the clustering radius, so as to obtain a plurality of clusters, the clustering number of the clusters, the clustering width of the clusters, and the clustering distance between the clusters. Cluster spacing refers to the spacing between a cluster and the nearest neighbor cluster.
In step 704 of some embodiments, when the number of clusters of the cluster body is less than the number of tray piers corresponding to the tray foot piers, the corresponding second point cloud data does not meet the specification of a tray to be taken, in which case, the corresponding second point cloud data cannot be processed as the point cloud data corresponding to the tray to be taken. Therefore, only when the clustering number of the clustering bodies is larger than or equal to the number of the tray piers corresponding to the tray foot piers, namely, only when the second point cloud data is confirmed to be the point cloud data corresponding to the tray to be taken, the subsequent judgment can be carried out.
The foot pier width threshold value refers to the actual pier width of the tray foot pier, and when the difference value between the cluster width and the foot pier width threshold value is smaller than the width difference threshold value, the cluster width is in accordance with the foot pier width threshold value. The width difference threshold is preset and can be modified according to actual conditions. For example, the width difference threshold may be selected to be 0.1cm.
When the difference value between the clustering distance between the nearest neighbor clustering bodies and the distance between the actual tray foot piers is smaller than the clamping plate error threshold value, the clustering distance is indicated to accord with the clamping plate error threshold value, and the clamping plate error threshold value is preset and can be modified according to the actual situation. For example, the threshold of the card error may be selected to be 0.1cm in this embodiment.
And when the clustering quantity of the clustering bodies is larger than or equal to the quantity of the tray piers corresponding to the tray foot piers, the clustering width accords with a foot pier width threshold value, and the clustering distance accords with a pallet error threshold value, the clustering result corresponding to the second point cloud data is described and can be used for representing that the second point cloud data comprises all tray foot piers.
Referring to fig. 8, in some embodiments, the method for identifying the pose of a tightly placed pallet includes the following steps:
And 804, filtering the original 3D point cloud data according to a preset target position to obtain first point cloud data.
And 808, projecting the target points to obtain a primary projection image, and performing gray level processing on the primary projection image to obtain an original tray image.
And step 822, filtering the original 3D point cloud data according to the tray identification area to obtain filtered point cloud data.
And step 824, fitting the filtered point cloud data to obtain target point cloud data corresponding to the tray to be taken.
And 830, if the clustering result corresponding to the second point cloud data represents that the second point cloud data comprises all tray foot piers, acquiring coordinates corresponding to each point cloud in the second point cloud data to obtain foot pier coordinate data.
And step 832, calculating the pose information of the tray to be taken according to the foot pier coordinate data.
Specifically, please refer to the embodiments in fig. 1 to 7 for the specific embodiments of steps 802 to 830, which are not described in detail herein.
It should be understood that, although the steps in the flowcharts related to the embodiments are shown in sequence as indicated by the arrows, the steps are not necessarily executed in sequence as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least a part of the steps in the flowcharts related to the above embodiments may include multiple steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of performing the steps or stages is not necessarily sequential, but may be performed alternately or alternately with other steps or at least a part of the steps or stages in other steps.
Based on the same inventive concept, the embodiment of the application also provides a compact-placement tray pose recognition device for realizing the compact-placement tray pose recognition method.
In one embodiment, as shown in fig. 9, there is provided a compact pallet pose recognition apparatus including: a point cloud acquisition module 902, an image projection module 904, a point cloud extraction module 906, and a pose calculation module 908, wherein:
a point cloud obtaining module 902, configured to obtain original 3D point cloud data of a target pickup area; wherein, the goods region is got to the target is placed and is waited to get the tray, and a plurality of trays of waiting to get are closely put, and the tray of waiting to get includes a plurality of tray foot mounds.
And an image projection module 904, configured to select multiple target points from the original 3D point cloud data, and project the multiple target points to obtain an original tray image.
And a point cloud extraction module 906, configured to extract target point cloud data corresponding to the to-be-taken tray from the original 3D point cloud data according to the original tray image.
And a pose calculation module 908 for calculating pose information of the pallet to be taken according to the target point cloud data.
In some embodiments, the image projection module 904 includes a preliminary filtering sub-module and a voxel filtering sub-module:
and the primary filtering submodule is used for filtering the original 3D point cloud data according to a preset target position to obtain first point cloud data.
And the voxel filtering submodule is used for carrying out voxel filtering on the first point cloud data to obtain a plurality of target points.
In some embodiments, point cloud extraction module 906 includes an image identification sub-module and a filtering sub-module, wherein:
the image recognition submodule is used for carrying out image recognition processing on the original tray image to obtain a tray recognition area; the tray identification area is used for representing an area of the tray to be taken in the original tray image; .
And the filtering submodule is used for filtering the original 3D point cloud data according to the tray identification area to obtain target point cloud data corresponding to the tray to be taken.
In some embodiments, the image recognition sub-module comprises an object recognition unit, a location information calculation unit, and a region filtering unit, wherein:
the target recognition unit is used for carrying out target recognition on the original tray image according to a preset target recognition algorithm to obtain a plurality of target recognition areas; wherein each target recognition area includes image position information of the tray in the original tray image.
And the position information calculating unit is used for calculating and obtaining the central position information corresponding to each target identification area according to the image position information.
The area screening unit is used for screening each target identification area according to preset target position and center position information to obtain a tray identification area; wherein, the tray identification area corresponds to the tray to be taken.
In some embodiments, the target identification unit comprises a grid partitioning subunit, a target detection subunit, an overlap computation subunit, and an edge region screening subunit, wherein:
and the grid division subunit is used for carrying out grid division on the original tray image to obtain a plurality of grid units.
And the target detection subunit is used for carrying out target detection on each grid unit to obtain a plurality of target detection areas corresponding to each grid unit.
And the overlap degree calculation subunit is used for calculating the boundary overlap degree corresponding to each target detection area according to the plurality of target detection areas corresponding to each grid unit and a preset target mark area.
And the edge area screening subunit is used for screening the target detection area according to the overlapping degree threshold and the boundary overlapping degree to obtain a target identification area.
In some embodiments, the pose calculation module comprises a point cloud segmentation sub-module, a clustering sub-module, a coordinate acquisition sub-module, and a pose calculation sub-module, wherein:
the point cloud segmentation submodule is used for segmenting the target point cloud data into a plurality of second point cloud data; each second point cloud data comprises a plurality of point clouds.
The clustering submodule is used for carrying out Euclidean clustering on each second point cloud data according to a preset clustering radius to obtain a clustering result corresponding to each second point cloud data; and the clustering result is used for representing whether the second point cloud data comprises all tray foot piers.
And the coordinate obtaining sub-module is used for obtaining the coordinate corresponding to each point cloud in the second point cloud data to obtain the foot pier coordinate data if the clustering result corresponding to the second point cloud data represents that the second point cloud data comprises all tray foot piers.
And the pose calculation submodule is used for calculating pose information of the tray to be taken according to the foot pier coordinate data.
In some embodiments, the clustering submodule includes an euclidean clustering unit and a determining unit, wherein:
and the Euclidean clustering unit is used for carrying out Euclidean clustering on each second point cloud data according to the clustering radius to obtain the clustering number, the clustering width and the clustering distance of the clustering bodies corresponding to each second point cloud data.
And the judging unit is used for representing that the second point cloud data comprises all the tray foot piers if the clustering number of the clustering body is greater than or equal to the number of the tray piers corresponding to the tray foot piers, the difference value between the clustering width and the foot pier width threshold value is smaller than the width difference threshold value, and the clustering distance is smaller than the clamping plate error threshold value.
All modules in the tray pose recognition device which is placed closely can be realized by software, hardware and a combination of the software and the hardware. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as shown in fig. 10. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program to run on the non-volatile storage medium. The communication interface of the computer device is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be realized through WIFI, a mobile cellular network, NFC (near field communication) or other technologies. The computer program is executed by a processor to implement a compact pose tray pose identification method. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
Those skilled in the art will appreciate that the architecture shown in fig. 10 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, there is provided a computer device comprising a memory in which is stored a computer program and a processor which, when executing the computer program, implements the method of compact pose tray pose identification in the embodiments of the first aspect.
In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which, when being executed by a processor, implements the method for identifying a pose of a closely-packed tray in the embodiments of the first aspect.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, database, or other medium used in the embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high-density embedded nonvolatile Memory, resistive Random Access Memory (ReRAM), magnetic Random Access Memory (MRAM), ferroelectric Random Access Memory (FRAM), phase Change Memory (PCM), graphene Memory, and the like. Volatile Memory can include Random Access Memory (RAM), external cache Memory, and the like. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM), among others. The databases involved in the embodiments provided herein may include at least one of relational and non-relational databases. The non-relational database may include, but is not limited to, a block chain based distributed database, and the like. The processors referred to in the embodiments provided herein may be general purpose processors, central processing units, graphics processors, digital signal processors, programmable logic devices, quantum computing based data processing logic devices, etc., without limitation.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present application. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.
Claims (10)
1. A method for identifying the pose of a tray placed closely, which is characterized by comprising the following steps:
acquiring original 3D point cloud data of a target goods taking area; the target goods taking area is used for placing trays to be taken, a plurality of trays to be taken are placed closely, and the trays to be taken comprise a plurality of tray foot piers;
selecting a plurality of target points from the original 3D point cloud data, and projecting the target points to obtain an original tray image;
extracting target point cloud data corresponding to the tray to be taken from original 3D point cloud data according to the original tray image;
and calculating the pose information of the tray to be taken according to the target point cloud data.
2. The method of claim 1, wherein selecting a plurality of target points from the raw 3D point cloud data comprises:
filtering the original 3D point cloud data according to a preset target position to obtain first point cloud data;
and carrying out voxel filtering on the first point cloud data to obtain a plurality of target points.
3. The method according to claim 1, wherein the extracting target point cloud data corresponding to the tray to be taken from original 3D point cloud data according to the original tray image comprises:
carrying out image recognition processing on the original tray image to obtain a tray recognition area; the tray identification area is used for representing an area of the tray to be taken in the original tray image;
filtering the original 3D point cloud data according to the tray identification area to obtain filtered point cloud data;
and fitting the filtered point cloud data to obtain target point cloud data corresponding to the tray to be taken.
4. The method according to claim 3, wherein the performing image recognition processing on the original tray image to obtain a tray recognition area comprises:
performing target recognition on the original tray image according to a preset target recognition algorithm to obtain a plurality of target recognition areas; wherein each target identification area comprises image position information of a tray in an original tray image;
calculating to obtain central position information corresponding to each target identification area according to the image position information;
screening each target identification area according to a preset target position and the central position information to obtain a tray identification area; wherein the tray identification area corresponds to the tray to be taken.
5. The method according to claim 4, wherein the performing target recognition on the original tray image according to a preset target recognition algorithm to obtain a plurality of target recognition areas comprises:
carrying out grid division on the original tray image to obtain a plurality of grid units;
carrying out target detection on each grid unit to obtain a plurality of target detection areas corresponding to each grid unit;
calculating to obtain the boundary overlapping degree corresponding to each target detection area according to the plurality of target detection areas corresponding to each grid unit and a preset target mark area;
and screening the target detection area according to the overlapping degree threshold and the boundary overlapping degree to obtain the target identification area.
6. The method according to any one of claims 1 to 5, wherein the calculating the pose information of the tray to be taken according to the target point cloud data comprises:
dividing the target point cloud data into a plurality of second point cloud data; wherein each of the second point cloud data comprises a plurality of point clouds;
ordering the second point cloud data to obtain ordered second point cloud data, and performing Euclidean clustering on each ordered second point cloud data according to a preset clustering radius to obtain a clustering result corresponding to each second point cloud data; the clustering result is used for representing whether the second point cloud data comprises all the tray foot piers;
if the clustering result corresponding to the second point cloud data represents that the second point cloud data comprises all tray foot piers, acquiring coordinates corresponding to each point cloud in the second point cloud data to obtain foot pier coordinate data;
and calculating the pose information of the tray to be taken according to the foot pier coordinate data.
7. The method of claim 6, wherein the performing Euclidean clustering on each second point cloud data according to a preset clustering radius to obtain a clustering result corresponding to each second point cloud data comprises:
performing Euclidean clustering on each second point cloud data according to the clustering radius to obtain the clustering number, the clustering width and the clustering distance of clustering bodies corresponding to each second point cloud data;
and if the clustering number of the clustering bodies is greater than or equal to the number of the tray piers corresponding to the tray pier, the difference value between the clustering width and the pier width threshold value is smaller than a width difference threshold value, and the clustering distance is smaller than a clamping plate error threshold value, the clustering result represents that the second point cloud data comprises all the tray piers.
8. A compact pallet pose recognition apparatus, the apparatus comprising:
the point cloud acquisition module is used for acquiring original 3D point cloud data of the target goods taking area; the target goods taking area is used for placing trays to be taken, a plurality of trays to be taken are placed closely, and the trays to be taken comprise a plurality of tray foot piers;
the image projection module is used for selecting a plurality of target points from the original 3D point cloud data and projecting the target points to obtain an original tray image;
the point cloud extraction module is used for extracting target point cloud data corresponding to the tray to be taken from original 3D point cloud data according to the original tray image;
and the pose calculation module is used for calculating pose information of the tray to be taken according to the target point cloud data.
9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the method of any of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211239101.5A CN115546300A (en) | 2022-10-11 | 2022-10-11 | Method and device for identifying pose of tray placed tightly, computer equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211239101.5A CN115546300A (en) | 2022-10-11 | 2022-10-11 | Method and device for identifying pose of tray placed tightly, computer equipment and medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115546300A true CN115546300A (en) | 2022-12-30 |
Family
ID=84734004
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211239101.5A Pending CN115546300A (en) | 2022-10-11 | 2022-10-11 | Method and device for identifying pose of tray placed tightly, computer equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115546300A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116342695A (en) * | 2023-05-29 | 2023-06-27 | 未来机器人（深圳）有限公司 | Unmanned forklift truck goods placing detection method and device, unmanned forklift truck and storage medium |
| CN116342858A (en) * | 2023-05-29 | 2023-06-27 | 未来机器人（深圳）有限公司 | Object detection method, device, electronic equipment and storage medium |
-
2022
- 2022-10-11 CN CN202211239101.5A patent/CN115546300A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116342695A (en) * | 2023-05-29 | 2023-06-27 | 未来机器人（深圳）有限公司 | Unmanned forklift truck goods placing detection method and device, unmanned forklift truck and storage medium |
| CN116342858A (en) * | 2023-05-29 | 2023-06-27 | 未来机器人（深圳）有限公司 | Object detection method, device, electronic equipment and storage medium |
| CN116342858B (en) * | 2023-05-29 | 2023-08-25 | 未来机器人（深圳）有限公司 | Object detection method, device, electronic equipment and storage medium |
| CN116342695B (en) * | 2023-05-29 | 2023-08-25 | 未来机器人（深圳）有限公司 | Unmanned forklift truck goods placing detection method and device, unmanned forklift truck and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115546300A (en) | Method and device for identifying pose of tray placed tightly, computer equipment and medium | |
| CN109844807B (en) | Method, system and apparatus for segmenting and sizing objects | |
| JP7478363B2 (en) | Forklift loading method, computer device and storage medium | |
| CN113191174A (en) | Article positioning method and device, robot and computer readable storage medium | |
| JP2021504262A (en) | Methods and system backgrounds for generating landing solutions for containers on landing | |
| CN114862301A (en) | Tray forklift AGV automatic loading method based on two-dimensional code auxiliary positioning | |
| CN116309882A (en) | Tray detection and positioning method and system for unmanned forklift application | |
| CN114332219B (en) | Tray positioning method and device based on three-dimensional point cloud processing | |
| CN115546202A (en) | Tray detection and positioning method for unmanned forklift | |
| CN114638974A (en) | Target object identification method, system, medium and electronic terminal | |
| CN114648233A (en) | Dynamic station cargo carrying method and system | |
| CN116038715B (en) | Box taking method, device, robot and storage medium | |
| CN114202548A (en) | Forklift pallet positioning method and device, storage medium and electronic equipment | |
| JP7352706B2 (en) | Cargo counting method and device, computer device, and storage medium | |
| CN116957459A (en) | Intelligent matching method, system, equipment and storage medium based on freight scene | |
| CN115100271A (en) | Method and device for detecting goods taking height, computer equipment and storage medium | |
| CN113978987B (en) | Pallet object packaging and picking method, device, equipment and medium | |
| CN114758163B (en) | Forklift movement control method and device, electronic equipment and storage medium | |
| CN115856923A (en) | Measuring method, device, equipment and storage medium for unloading of mine truck | |
| CN116071547A (en) | Tray pose detection method and device, equipment and storage medium | |
| CN115062737A (en) | Method, device and equipment for acquiring cargo pose based on 2d camera and storage medium | |
| CN115330854A (en) | Cargo management system and cargo management method | |
| CN115100283A (en) | Goods placement control method and device, computer equipment and storage medium | |
| CN114819821A (en) | Goods warehouse-out checking method and device, computer equipment and storage medium | |
| CN116433742A (en) | Method and device for judging rule violation of transport vehicle, electronic equipment and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |