How to determine whether stacking is possible by measuring the top surface
The method automates stackability determination by measuring the top surface of returnable boxes using a 3D measuring device, addressing inefficiencies in existing point cloud data methods to ensure precise and efficient stacking decisions.
Patent Information
- Application Number
- JP2021202334
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2021-12-14
- Publication Date
- 2024-11-13
- Estimated Expiration
- 2041-12-14
AI Technical Summary
Existing 3D sensor-based point cloud data measurement methods are inefficient for acquiring necessary height data and require manual pattern input, hindering automation in determining stackability of returnable boxes.
A method involving top surface measurement of the topmost box using a 3D measuring device, including input, determination processing, and output steps, which includes plane correction, noise removal, top surface data extraction, corner detection, and center of gravity calculation to determine stackability.
Enables highly accurate and automated determination of stackability, optimizing the placement of returnable boxes by measuring the top surface with high precision and efficiency.
Smart Images

Figure 0007582171000001 
Figure 0007582171000002 
Figure 0007582171000003
Abstract
Description
[Technical field]
[0001] The present invention relates to a ceiling surface measurement method. [Background technology]
[0002] Conventionally, when transferring workpieces, it is necessary to optimally position a returnable box containing the workpieces on a pallet and then stack another pallet with a similar returnable box on it, so it is necessary to measure the top surface of the returnable box placed on the pallet, and measurement techniques for this purpose are known.
[0003] For example, Patent Document 1 discloses a system for measuring the height of a workpiece to be transferred, which acquires three-dimensional point cloud information of a plurality of loaded workpieces and accurately measures the height of the workpiece to be transferred. [Prior art documents] [Patent documents]
[0004] [Patent Document 1] Patent Publication No. 2021-36204 Summary of the Invention [Problem to be solved by the invention]
[0005] In the case of point cloud data measurement using a 3D sensor as in Patent Document 1, it is not possible to efficiently obtain only the necessary height data. Also, when trying to pattern the data capture range, the number of patterns becomes enormous, requiring manual pattern input by an operator, making full automation difficult.
[0006] The present invention was created in consideration of these problems, and its purpose is to provide a top surface measurement method that enables highly accurate measurement of the top surface of a returnable container and enables the automation of determination of whether or not the container can be stacked. [Means for solving the problem]
[0007] The present invention provides a stackability determination method (1) by measuring the top surface of a pallet, which, when rectangular parallelepiped boxes of the same height are stacked in one or more layers in four or more places on the pallet, measures the top surface data of the topmost box among these boxes and determines whether another pallet can be stacked on top of the topmost box on the pallet, and includes an input step (11), a determination processing step (12), and a determination result output step (13). The input step inputs point cloud data of the entire boxes measured by a 3D measuring device (3). The determination processing step extracts the top surface data of the topmost box from the point cloud data and performs calculation processing to determine whether the boxes can be stacked. The output step outputs the determination result. The determination process steps include a top surface Z value detection step (123) for detecting Z values constituting the top surface of the top box from the point cloud data, a top surface data extraction step (124) for extracting top surface data including the Z values constituting the top surface, a corner detection step (125) for detecting corners of a polygon on the XY plane formed by the top surface based on the X and Y values of the top surface data, a center of gravity calculation step (128) for calculating the coordinates of the center of gravity of the polygon and calculating the distance from the coordinates of the center of gravity of the pallet measured in advance, and a determination step (129) for determining whether or not stacking is possible. In the determination step, if the distance of the coordinates of the center of gravity of the polygon from the coordinates of the center of gravity of the pallet exceeds a preset reference value in the center of gravity calculation step, it is determined that stacking is not possible.
[0008] This enables highly accurate measurement of the top surface of returnable containers, automating the decision as to whether or not they can be stacked. [Brief description of the drawings]
[0009] [Figure 1] FIG. [Diagram 2] A perspective view of all the boxes on the pallet. [Diagram 3] Front view of Figure 2. [Figure 4] FIG. [Diagram 5] 3 illustrates an example of a device configuration according to an embodiment. [Figure 6] The relative positions of the 3D measuring device and the pallet. [Figure 7] 1 is a flow diagram of one embodiment. [Figure 8] 1 is a flow chart of determination processing steps according to an embodiment. [Figure 9] Another perspective view of the entire box on the pallet. [Figure 10] Front view of Figure 8. [Figure 11] An example of the top surface detected in the top surface Z value detection step. [Figure 12] An example of the corner detection step. [Figure 13] Another example of the corner detection step. [Figure 14] The position of the palette relative to the rectangular area. [Figure 15] An example of edge measurement steps. [Figure 16] Example of area measurement steps. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0010] An embodiment of the present invention will be described below with reference to the drawings. Figures 1 to 4 show an application of the present invention. This embodiment assumes a situation in which a plurality of standard returnable boxes containing parts or products are arranged on a pallet P as shown in Figures 2 and 3, as shown in Figures 1, and these pallets P are then stacked on top of another similar pallet P, each with its own returnable box.
[0011] In this embodiment, the space in which the pallet P and returnable containers exist is a 3D coordinate system in which a plane parallel to the floor is the XY plane and the positive direction of the Z axis is the upward direction. The shape of the pallet P on the XY plane is approximately square. The shape of the top surface of an actual returnable container is as shown in Figure 4, but in Figures 1 to 3, 6, 9, 11 to 13, 15, and 16 it is abbreviated as a solid rectangular parallelepiped.
[0012] The present invention is a stackability determination method 1 by measuring the top surface when rectangular parallelepiped boxes of the same size are stacked in one or more layers in four or more places on a pallet, measuring the top surface data of the topmost box among these boxes and determining whether other pallets can be stacked on top of the topmost box on the pallet, the method including an input step 11 of inputting point cloud data of all the boxes measured by a 3D measurement device 3, a determination processing step 12 of extracting the top surface data of the topmost box from the point cloud data and performing calculation processing to determine whether the boxes can be stacked, and an output step 13 of the determination result. Hereinafter, "stackability determination method 1 by measuring the top surface" will be abbreviated as "determination method 1".
[0013] (One embodiment) FIG. 5 shows an example of an equipment configuration for implementing one embodiment of the present invention. As shown in FIG. 5, the determination method 1 is implemented by equipment including a computer 2 for performing data processing and a 3D measurement device 3. FIG. 6 shows the positional relationship between the 3D measurement device 3 and the entire box Bx on the pallet P that is the measurement target. As shown in FIG. 6, the 3D measurement device 3 is installed in a position that overlooks the measurement target from above, and accurately measures the distance between the 3D measurement device 3 and the measurement target (the entire box Bx). Note that in this specification, the "top surface" refers to the highest plane among the "top surfaces" of the entire box Bx that are approximately the same height and parallel to the XY plane. Therefore, it does not include planes that are lower by one or more times the height of the box B.
[0014] A flow of one embodiment is shown in Figure 7. The judgment method 1 of the present invention includes an input step 11, a judgment processing step 12, and an output step 13. In the input step 11, point cloud data Do of the entire box Bx measured by the 3D measurement device 3 is input. In the judgment processing step 12, top surface data Dt is extracted from the point cloud data Do and a calculation process is performed to judge whether the boxes can be stacked. In the output step 13, the judgment result is output. In this embodiment, the rectangular parallelepiped box B of the same size is a standard shaped so-called returnable box.
[0015] 8 shows a flow of the determination processing step 12 according to one embodiment. In this embodiment, the determination processing step 12 includes a plane correction step 121, a noise removal step 122, a top surface Z value detection step 123, a top surface data extraction step 124, a corner detection step 125, a side measurement step 126, an area measurement step 127, a center of gravity calculation step 128, and a determination step 129.
[0016] In a plane correction step 121, the inclination of the point cloud data Do is corrected based on pre-measured data on the inclination of the floor surface on which the pallet P is placed. In a noise removal step 122, the point cloud data Do is smoothed to remove unnecessary data (noise) such as fixing strings attached to packages in returnable boxes, and to eliminate abnormal values such as outliers. Any smoothing process can be used for this, but a median filter is used in this embodiment.
[0017] Next, in order to make the determination process more efficient, only the top surface data Dt necessary and sufficient for top surface measurement is extracted from the point cloud data Do, which has a large amount of data. In the top surface Z value detection step 123, the Z value constituting the top surface of the topmost box is detected from the point cloud data Do. Any detection method can be used, but in this embodiment, a histogram is created for the height (Z value), and data that occurs with a frequency where the height exceeds a preset reference value is detected, and a Z value that falls within a range that satisfies a preset determination condition is identified. In the following top surface data extraction step 124, the point cloud data to which the Z value detected in the top surface Z value detection step 123 belongs, i.e., the top surface data Dt, is extracted.
[0018] Since there can be a wide variety of top surface shapes, some shapes may not be able to be stacked. For example, even if there is a step due to an empty space, there are cases where stacking another pallet P is possible without any problems, as in Figures 2-3, and there are also cases where there is a risk of collapse when stacked, as in Figures 9-10, due to the obvious lack of stability. The shapes of the top surface (shaded area) in each case are shown in Figures 11(a) and (b). For this reason, it is inefficient and unrealistic to perform pattern recognition on all variations in top surface shapes.
[0019] For this reason, the determination method 1 of this embodiment employs the following steps: (1) The top surface shape is identified in a corner detection step 125. (2) The characteristics of the top surface shape are grasped in a side measurement step 126 and an area measurement step 127. (3) If the top surface shape meets the criteria, the center of gravity of the entire box Bx is measured in a center of gravity calculation step 127, and stackability is determined based on the deviation between that center of gravity and the center of gravity of the pallet P.
[0020] 12-13 show an outline of the corner detection step 125. The corner detection step 125 is a step for identifying the corners of the polygon formed by the top surface shape based on the X and Y values of the top surface data Dt. Closely adjacent and distant points in the top surface data Dt are connected to each other with line segments, and a polygon Q containing them and its corners Qc are identified. The corners Qc identified at this time include corners Qc that are convex on the outside of the polygon (white circles in FIGS. 12-13) and corners Qc that are convex on the inside of the polygon (black circles in FIGS. 12-13). When identifying the shape of each polygon from these corner regions Qc, either a non-convex hull pattern that identifies the rough shape as it is, or a convex hull pattern that identifies the polygon without any recesses is selected. In the case of a convex hull pattern, the polygon is formed only by the corners Qc (white circles) that are convex on the outside.
[0021] 14 shows an example of the rectangular region R used in the side measurement step 126 and the area measurement step 127. In the side measurement step 126 and the area measurement step 127, four rectangular regions R are mechanically set along each side of the pallet P, and the space on the pallet P is divided as shown in FIG.
[0022] An example of the side measurement step 126 is shown in Fig. 15. In the side measurement step 126, the lengths of the sides of each rectangular region R in the region that overlaps with the polygon Q are measured. In the example of Fig. 15, the lengths L1 to L4 are measured, and if any one of them does not reach the preset length reference value, it is determined that the stacking is not possible.
[0023] An example of the area measurement step 127 is shown in Fig. 16. In the area measurement step 127, the area of each rectangular region R that does not overlap with the polygon Q is measured. In the example of Fig. 16, the areas of the spaces W1H1 and W2H2 where the polygon Q does not exist are measured in each of the regions W1H1, W2H2, W3H3, and W4H4. If even one point exceeds a preset reference area value, it is determined that stacking is not possible.
[0024] If stacking is not determined to be impossible in either the side measurement step 126 or the area measurement step 127, it is determined that stacking is possible at least on the top surface shape. These determination results are provided to a determination step described later.
[0025] In the center of gravity calculation step 128, the center of gravity coordinate QG is calculated for the polygon Q identified in the corner detection step 125. At this time, whether the polygon Q is identified by a non-convex hull pattern or by a convex hull pattern, the center of gravity calculation is performed for the polygon Q as a solid polygon (FIGS. 12 to 13). Then, the distance GS = |PG-QG| between the calculated center of gravity coordinate QG and the center of gravity coordinate PG of the pallet P measured in advance is calculated. If the distance GS exceeds a preset reference value, it is determined that stacking is not possible. If it is not determined that stacking is not possible, it is determined that stacking is possible based on the balance of the centers of gravity of the entire box Bx and the pallet. The result of the determination is provided to the determination step described later.
[0026] In decision step 129, it can be determined whether or not the objects can be stacked based on the results of side measurement step 126, area measurement step 127, or center of gravity calculation step 128. In this embodiment, however, all of these results are taken into consideration, and if any of the results do not satisfy the criteria, it is determined that the objects cannot be stacked.
[0027] [Effects of one embodiment] By adopting this embodiment, it is possible to measure the top surface of returnable containers with high accuracy and efficiency using only a 3D measuring device and a computer, which helps to automate and speed up the stackability determination process, and also provides information that is useful for planned optimal placement. Therefore, it is useful for fully automating and speeding up the entire system.
[0028] (Other embodiments) The present invention can be applied not only to returnable containers as shown in FIG. 4, but also to solid rectangular parallelepipeds in the same manner.
[0029] The plane correction step may be omitted if it is not necessary. The noise removal step may also be omitted after the first measurement depending on the situation. The algorithm used in the corner detection step may be other than that shown in the above embodiment, and any algorithm may be selected. Neither the side measurement step nor the area measurement step may be performed depending on the necessity, or only one of them may be performed.
[0030] In the above embodiment, the center of gravity calculation step uses the values of the polygon detected in the corner detection step, but instead of the polygon, the point cloud data of the entire box initially measured by the 3D measuring device may be directly processed, or the top surface data extracted in the top surface data extraction step prior to processing in the subsequent corner detection step may be processed.
[0031] Although the present invention is specified as a method invention, it may also be embodied as a system or computer program. In the case of a computer program, it may include a computer readable storage medium having computer readable program instructions for causing a processor to carry out the invention. The computer readable storage medium may be a tangible device capable of storing instructions for use by an instruction execution device.
[0032] As described above, the present invention is not limited to the above-described embodiment, and can be embodied in various forms without departing from the spirit of the present invention. [Explanation of symbols]
[0033] 1. How to determine whether pallets can be stacked 11 Input Step 12. Judgment process step 121 Plane correction step 122 Noise removal step 123 Top surface Z value detection step 124 Top surface data extraction step 125 ···Angle detection steps 126...Side measurement step 127 Area measurement step 128 Center of gravity calculation steps 129 Judgment step 13 Output Step 3. 3D Measuring Device
Claims
1. A method (1) for determining whether or not a pallet can be stacked on top of a top-level box on a pallet, the method measuring the top surface data of the topmost box among the boxes when the boxes are stacked in one or more layers in four or more places on the pallet, and determining whether or not another pallet can be stacked on top of the topmost box on the pallet, comprising: an input step (11) of inputting point cloud data of the entire box measured by the 3D measuring device (3); A determination processing step (12) of extracting top surface data of the topmost box from the point cloud data and performing calculation processing to determine whether the boxes can be stacked; A step (13) of outputting the determination result; Including, The determination process step includes: a top surface Z value detection step (123) for detecting Z values constituting the top surface of the topmost box from the point cloud data; A top surface data extraction step (124) of extracting top surface data including Z values constituting the top surface; a corner detection step (125) for detecting corners of a polygon on an XY plane formed by the top surface based on the X and Y values of the top surface data; a center of gravity calculation step (128) for calculating the coordinates of the center of gravity of the polygon and calculating the distance from the coordinates of the center of gravity of a previously measured pallet; A determination step (129) for determining whether stacking is possible; Including, The method for determining whether or not stacking is possible by measuring the top surface, wherein the determination step determines that stacking is not possible if the distance between the center of gravity coordinates of the polygon and the center of gravity coordinates of the pallet exceeds a predetermined reference value in the center of gravity calculation step.
2. The determination process step further includes a side measurement step (126) of measuring the length of a side of an area where each rectangular area overlaps with the polygon when the rectangular area is superimposed on the polygon and the rectangular area is set to divide the space on the pallet along each side of the pallet, The method for determining whether or not the objects can be stacked by measuring the top surface as described in claim 1, wherein the judgment step determines that the objects cannot be stacked if the length of the side of the area overlapping with the polygon in the side measurement step does not reach a predetermined reference value at any point.
3. The determination process step further includes an area measurement step (127) of measuring an area of each rectangular area that is not overlapped with the polygon when the rectangular area is overlapped with the polygon, the area measurement step being performed to determine an area of each rectangular area that is not overlapped with the polygon when the rectangular area is overlapped with the polygon.
2. The method for determining whether the objects can be stacked by measuring the top surface according to claim 1, wherein the determination step determines that the objects cannot be stacked if the area of any one area not overlapping with the polygon in the area measurement step exceeds a preset reference value.
4. The determination process step includes: A plane correction step (121) of correcting the inclination of the point cloud data based on the inclination of the floor surface measured in advance, or a noise removal step (122) for smoothing the point cloud data; The method for determining whether stacking is possible by measuring the top surface according to claim 1 , further comprising:
Citation Information
Patent Citations
Method and apparatus for palletizing package of random size and weight
JP2003335417A
Forklift
JP2018062421A
Recessed part detection device, carrier device, and recessed part detection device
JP2018179599A
Transfer object work-piece height measuring system
JP2021036204A
Article handling device
JP2022011857A
