Cargo amount calculation method, cargo identification method, and cargo amount calculation apparatus

Abstract

To provide a cargo amount calculation method, a cargo identification method, and a cargo amount calculation apparatus which can grasp a cargo amount of a truck with high reliability.SOLUTION: The present invention is directed to a cargo amount calculation method of calculating a cargo amount in a loading part 94 of a truck 92. The method of calculating a cargo amount of the truck 92 has a step of acquiring a first loading image obtained by capturing the loading part 94 from one side in a vehicle width direction of the truck 92 and from an outside thereof, a step of comparing, in the first loading image, the loading part 94 with the lowermost pallet 95 in a cargo bulk 96 where a cargo 90 is loaded, a step of, if the pallet 95 is located within a predetermined region of the loading part 94, identifying the cargo bulk 96 as a cargo amount calculation target, and a step of calculating the amount of the cargo 90 included in the cargo bulk 96.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to a technique for grasping the amount of load actually loaded on the loading part of a truck.

Background Art

[0002] Trucks for collecting and delivering various goods are provided with a loading part for loading the goods.

[0003] In recent years, attempts have been made to grasp the amount of load actually loaded on the loading part of a truck. This is for managing the collection and delivery schedule of goods by the truck, for managing whether an appropriate amount of goods is loaded on the loading part of the truck, or for managing whether the goods are loaded at an appropriate position in the loading part.

[0004] As one method for reducing the involvement of workers in load amount management, there is a method of imaging the loading part of a truck and calculating the load amount and free space of the loading part based on the obtained loading image. Patent Document 1 and Patent Document 2 introduce techniques for imaging the loading part of a truck.

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0006] Patent Document 1 introduces a technique of imaging the loading part with a camera installed on the wing of a truck and calculating the load amount loaded on the loading part based on the obtained loading image. In addition, Patent Document 2 introduces a technique of imaging a loading part of a truck and a container, that is, the truck, with a camera.

[0007] If the loading part is imaged using a camera as in the techniques introduced in Patent Document 1 and Patent Document 2 described above, it is considered that the load amount actually loaded on the loading part of the truck can be grasped based on the obtained image.

[0008] However, in reality, it is difficult to accurately grasp the load amount actually loaded on the loading part simply by imaging the loading part using a camera. In particular, when automatically identifying the load amount based on an image of the loading part, it is very difficult to accurately grasp the load amount.

[0009] Therefore, in order to manage the collection and delivery schedules of goods by trucks, to manage whether an appropriate amount of goods is loaded on the loading part of the truck, or to manage whether goods are loaded at an appropriate position in the loading part, a technique capable of reliably grasping the load amount of the truck is desired.

[0010] The present invention has been made in view of the above circumstances, and an object to be solved is to provide a load amount calculation method, a cargo identification method, and a load amount calculation device capable of reliably grasping the load amount of a truck.

Means for Solving the Problems

[0011] The load amount calculation method of the present invention for solving the above problems is a load amount calculation method for calculating the load amount in the loading part of a truck, acquiring a first loading image obtained by imaging the loading part from one side in the vehicle width direction outside the vehicle with respect to the truck, in the first loading image, comparing the loading part with the lowermost pallet in a load stack on which goods are loaded, A method for calculating the load capacity of a truck, which, when the pallet is within a predetermined area of the loading section, identifies that the load mountain is a target for load capacity calculation and calculates the quantity of the goods contained in the load mountain.

[0012] Moreover, a method for identifying goods according to the present invention for solving the above problems is A method for identifying goods that identifies whether the goods loaded on the loading section of a truck are targets for load capacity calculation, acquires a first loading image obtained by imaging the loading section from the outside of the vehicle and from one side in the vehicle width direction with respect to the truck, in the first loading image, compares the loading section with the lowermost pallet in the load mountain on which goods are loaded, and, when the pallet is within a predetermined area of the loading section, identifies that the load mountain is a target for load capacity calculation. A method for identifying goods on a truck.

[0013] Furthermore, a load capacity calculation device according to the present invention for solving the above problems is a load capacity calculation device for calculating the load capacity in the loading section of a truck, a camera installed on the outside of the vehicle and on one side in the vehicle width direction with respect to the truck, which images the loading section to obtain a first loading image, in the first loading image, compares the loading section with the lowermost pallet in the load mountain on which goods are loaded, and when the pallet is within a predetermined area of the loading section, identifies that the load mountain is a target for load capacity calculation and calculates the quantity of the goods contained in the load mountain. A load capacity calculation device for a truck, comprising a load management element.

Advantages of the Invention

[0014] According to the load capacity calculation method, goods identification method, and load capacity calculation device of the present invention, it is possible to accurately grasp the load capacity of a truck.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Modes for Carrying Out the Invention

[0016] Hereinafter, the load calculation method, the cargo identification method, and the load calculation device of the present invention will be described with specific examples.

[0017] Unless otherwise specified hereinafter, the numerical range "x to y" described in this specification includes the lower limit x and the upper limit y in the range. And by arbitrarily combining these upper limit values, lower limit values, and the numerical values listed in the embodiments, a numerical range can be constituted. Further, a numerical value arbitrarily selected from within the numerical range can be used as the numerical values of the upper limit and the lower limit.

[0018] As described above, like the technology introduced in Patent Document 1 and the conventional technology introduced in Patent Document 2, it has been difficult to accurately grasp the actual load on the loading part by simply using a camera to image the loading part.

[0019] The inventor of the present invention obtained the idea that the above problem is closely related to the environment for imaging the loading part of the truck.

[0020] In order to image the loading section for the purpose of grasping the load quantity loaded on the loading section, it is preferable to perform imaging immediately before the truck loaded with goods on the loading section departs during cargo collection, or immediately before unloading the goods from the loading section of the truck during shipment. Considering this, as an environment for imaging the loading section, a cargo loading and unloading base for collecting and delivering goods is suitable.

[0021] However, the cargo loading and unloading base is a facility used by many trucks simultaneously. As shown in FIG. 1, there may be a case where a truck not subject to load quantity calculation parks near the truck subject to load quantity calculation. In this case, when trying to image the loading section at the loading and unloading base, a truck not subject to imaging overlaps with the imaging target truck on the front side in the imaging direction. As a result, as shown in FIG. 2, there is a possibility that the image obtained by imaging the loading section may include a truck not subject to load quantity calculation, in other words, the loading section of a truck not subject to imaging, and the goods on the loading section.

[0022] The inventor of the present invention noticed that when the loading section of a truck not subject to load quantity calculation is reflected in the image of the loading section, it becomes difficult to accurately distinguish between the goods loaded on the loading section of the truck subject to load quantity calculation and the goods loaded on the loading section of the truck not subject to load quantity calculation. And the inventor conceived that as a result, it also becomes difficult to accurately grasp the actual load quantity loaded on the loading section.

[0023] The inventor of the present invention further advanced the idea, aimed to accurately distinguish between the goods loaded on the loading section of the truck subject to load quantity calculation and the goods loaded on the loading section of the truck not subject to load quantity calculation when the loading section of a truck not subject to load quantity calculation is reflected in the image of the loading section, and conducted intensive studies. As a result, the present invention's cargo identification method, load quantity calculation method, and load quantity calculation device were invented.

[0024] In the luggage identification method of the present invention, a first loading image is acquired by imaging the loading part from one side outside the vehicle and in the vehicle width direction with respect to the truck. If there is luggage loaded on the loading part, in the first loading image, in addition to the loading part itself, the luggage loaded on the loading part is imaged in the state of a stack of luggage on the pallet.

[0025] In the luggage identification method of the present invention, it is identified whether the luggage shown in the above first loading image is the luggage loaded on the loading part of the truck whose load amount is to be calculated, or the luggage loaded on the loading part of the truck whose load amount is not to be calculated.

[0026] Specifically, in the luggage identification method of the present invention, first, in the above first loading image, the loading part and the lowermost pallet in the stack of luggage are compared. And when the pallet is within a predetermined area of the loading part, it is identified that the stack of luggage is the object for which the load amount is to be calculated. It can also be said that in the luggage identification method of the present invention, based on the positional relationship between the loading part in the first loading image and the lowermost pallet in the stack of luggage, it is identified whether the stack of luggage is the object for which the load amount is to be calculated.

[0027] The outer shape of the loading part has nothing to do with the presence or absence of luggage. Therefore, it can be said that it is appropriate to use the loading part itself as a criterion for identifying whether the luggage shown in the first loading image is the luggage loaded on the loading part of the truck whose load amount is to be calculated, or the luggage loaded on the loading part of the truck whose load amount is not to be calculated, that is, as an identification criterion in the first loading image.

[0028] On the other hand, the outer shapes of the luggage loaded on the loading part and the stack of luggage on which the luggage is loaded change three-dimensionally as the luggage is loaded and unloaded on the loading part. Therefore, it is difficult to say that it is appropriate to use the outer shape of the stack of luggage itself as an identification criterion in the first loading image.

[0029] Here, the lowermost pallet on the cargo mountain with the cargo loaded thereon only appears, disappears, or changes its two-dimensional position on the placement surface of the loading section as the cargo is loaded and unloaded onto and from the loading section. Therefore, the lowermost pallet on the said cargo mountain can be said to be useful as an identification criterion in the first loading image.

[0030] For this reason, in the cargo identification method of the present invention, the positional relationship between the loading section and the lowermost pallet on the cargo mountain is adopted as the identification criterion in the first loading image. When the lowermost pallet on the cargo mountain is within a predetermined area of the loading section, it is identified that the said cargo mountain is an object for calculating the cargo quantity. In other words, when the lowermost pallet on the cargo mountain is not within the predetermined area of the loading section, it is identified that the said cargo mountain is not an object for calculating the cargo quantity. Thereby, it is possible to accurately identify whether the cargo shown in the first loading image is the cargo loaded on the loading section of the truck that is an object for calculating the cargo quantity, or the cargo loaded on the loading section of the truck that is not an object for calculating the cargo quantity.

[0031] Also, according to the cargo quantity calculation method of the present invention, the quantity of the cargo included in the cargo mountain identified as an object for calculating the cargo quantity by the cargo identification method of the present invention is calculated. Thus, according to the cargo quantity calculation method of the present invention using the cargo identification method of the present invention, it is possible to accurately calculate the cargo quantity of the truck that is an object for calculating the cargo quantity.

[0032] The cargo quantity calculation device of the present invention is a device comprising a camera and a cargo quantity management element capable of implementing the above-described cargo quantity calculation method of the present invention. Thus, according to the cargo quantity calculation device of the present invention, it is possible to accurately calculate the cargo quantity of the truck that is an object for calculating the cargo quantity.

[0033] Hereinafter, the cargo identification method, the cargo quantity calculation method, and the cargo quantity calculation device of the present invention will be described for each of their components. In the case where it is simply referred to as the present invention without particular explanation, it shall be taken as comprehensively referring to the cargo identification method, the cargo quantity calculation method, and the cargo quantity calculation device of the present invention.

[0034] Hereinafter, if necessary, the truck whose load is to be calculated may be referred to as the target truck, and the loading part of the target truck may be referred to as the target loading part. Further, the pallet at the lowermost stage of the cargo mountain may be referred to as the lower pallet if necessary. For reference, the pallet referred to in the present invention means a loading platform for simultaneously transporting a plurality of packages using a mobile lifting device typified by a forklift.

[0035] In the present invention, first, a first loading image obtained by imaging the loading part of the target truck from the outside of the vehicle and one side in the vehicle width direction is acquired. One side in the vehicle width direction referred to here may be any specific one side in the vehicle width direction. For example, it may be the right side or the left side when the traveling direction of the target truck is defined as the front side, or a side including the right side or the left side, for example, the right side and the rear side, or the right side, the rear side, and the upper side, etc.

[0036] In the present invention, the camera for imaging the loading part is installed on the outside of the vehicle and on one side in the vehicle width direction with respect to the target truck in order to acquire the first loading image. The camera may be capable of imaging a moving image or a still image. That is, the first loading image captured by the camera may be a moving image or a still image.

[0037] Note that the loading part of the truck refers to the part of the truck on which packages are loaded, and can also be referred to as the cargo compartment of the truck.

[0038] In the present invention, the loading part in the first loading image and the pallet at the lowermost stage of the cargo mountain in the first loading image, that is, the lower pallet, are compared. Note that depending on the positional relationship between the target truck and the camera, it is conceivable that a plurality of loading parts are shown in the first loading image. In such a case, the largest loading part in the first loading image, that is, the loading part closest to the camera, may be defined as the target loading part.

[0039] The number of loads on the loading section varies according to the consolidation and delivery status of the target truck and can take various values. Therefore, the number of loads in the first loading image may be only one or may be multiple. Furthermore, there may be no load.

[0040] When there are multiple loads in the first loading image, for each load, its lower pallet may be compared with the target loading section.

[0041] As described above, by comparing the target loading section with the lower pallet, it is possible to determine whether the lower pallet is within a predetermined area of the target loading section. And when the lower pallet is within the predetermined area, it becomes possible to identify that the load including the lower pallet is a target for load calculation.

[0042] Here, even for loads with the same positional relationship with respect to the target loading section in the first loading image, depending on the range of the predetermined area as the identification criterion, there may be cases where they are identified as targets for load calculation and cases where they are not identified as targets for load calculation. Therefore, the predetermined area as the identification criterion may be appropriately set based on the matter of which loads with any positional relationship with respect to the target loading section are identified as targets for load calculation.

[0043] As an example, (a) in the first loading image, loads on the side closer to the reference on one side in the vehicle width direction, in other words, on the side closer to the camera, can be taken as targets for load calculation. According to (a) above, loads on the side farther from the reference with respect to the target loading section, in other words, on the side farther from the camera, can be excluded from the targets for load calculation.

[0044] In addition, when identifying targets for load calculation according to the above (a), it is more suitable to use the first loading image taken from an oblique direction including the one side rather than simply using the one taken from one side of the target truck to image the loading section. Specifically, it is the one side and the rear side, the one side and the front side, the one side and the rear side and the upper side, the one side and the front side and the upper side, etc.

[0045] In the above (a), it is preferable that one end side in the vehicle width direction on the floor surface of the target loading part is used as the first reference line, and a range where the distance from the first reference line is equal to or less than a predetermined value is set as a predetermined area.

[0046] For example, if a range where the distance from the above first reference line toward the other side in the vehicle width direction is equal to or less than the length of the target loading part in the vehicle width direction (in other words, the width of the target loading part) is set as the predetermined area, loads on the other side in the vehicle width direction than the target loading part can be excluded from the load amount calculation target, and other loads can be set as the load amount calculation target. Also, for example, if a range where the distance from the first reference line toward the other side in the vehicle width direction is equal to or less than half of the width of the target loading part is set as the predetermined area, only the loads on one side in the vehicle width direction in the target loading part can be set as the load amount calculation target.

[0047] As another example, in (b), in the first loading image, taking a predetermined position in the vehicle traveling direction as a reference, loads on the side closer to the reference can be cited as the load amount calculation target. In the above (b), loads on the side farther from the reference with respect to the target loading part, in other words, loads on the front side or the rear side in the vehicle traveling direction than the target loading part, can be excluded from the load amount calculation target.

[0048] In the above (b), (b-1) it is preferable that the end side on the rear side in the vehicle traveling direction on the floor surface of the target loading part is used as the rear reference line, and a range where the distance from the rear reference line toward the front side in the vehicle traveling direction is equal to or more than a predetermined value, or a range where the distance from the rear reference line toward the rear side in the vehicle traveling direction is equal to or less than a predetermined value is set within the predetermined area. Or, (b-2) it is also preferable that the end side on the front side in the vehicle traveling direction on the floor surface of the target loading part is used as the front reference line, and a range where the distance from the front reference line toward the rear side in the vehicle traveling direction is equal to or more than a predetermined value, or a range where the distance from the front reference line toward the front side in the vehicle traveling direction is equal to or less than a predetermined value is set within the predetermined area.

[0049] In (b-1), for example, if a range where the distance from the above-mentioned rear reference line toward the front side in the vehicle traveling direction exceeds 0 is defined as a predetermined area, loads located behind the target loading part in the vehicle traveling direction can be excluded, and other loads can be set as targets for load amount calculation.

[0050] In (b-2), for example, if a range where the distance from the above-mentioned front reference line toward the rear side in the vehicle traveling direction exceeds 0 is defined as a predetermined area, loads located in front of the target loading part in the vehicle traveling direction can be excluded, and other loads can be set as targets for load amount calculation.

[0051] Furthermore, in the above-mentioned (b-1), for example, if a range where the distance from the rear reference line toward the front side in the vehicle traveling direction is equal to or greater than half of the length of the target loading part in the vehicle traveling direction is defined as a predetermined area, loads located behind the target loading part in the vehicle traveling direction can be excluded, and loads located in front of the target loading part in the vehicle traveling direction can be set as targets for load amount calculation.

[0052] Similarly, in the above-mentioned (b-2), for example, if a range where the distance from the front reference line toward the rear side in the vehicle traveling direction is equal to or less than half of the length of the target loading part in the vehicle traveling direction is defined as a predetermined area, loads located behind the target loading part in the vehicle traveling direction can be excluded, and loads located in front of the target loading part in the vehicle traveling direction can be set as targets for load amount calculation.

[0053] When there are few loads on the loading part, for example, when the truck makes an emergency stop, the loads may move forward on the loading part in the vehicle traveling direction. If the moving distance of the loads is excessive, the loads may collide with the front wall of the loading part, and the goods contained in the loads may be damaged.

[0054] On the other hand, by arranging the loads on the front side of the loading part in the vehicle traveling direction, the movable distance of the loads moving forward in the vehicle traveling direction can be shortened. As a result, even when there are few loads on the loading part and the truck makes an emergency stop, excessive movement of the loads can be suppressed, and thus damage to the goods contained in the loads can be suppressed.

[0055] In the present invention, by setting the load mountain in the front side in the vehicle traveling direction in the target loading section as the load amount calculation target according to the above (b-1) or (b-2), when there are few or no load mountains that are the load amount calculation targets and there are load mountains that are not the load amount calculation targets, it is possible to alert the operator or give an instruction to rearrange the load mountains on the work robot. As a result, when there is a load mountain that is not the load amount calculation target in the target loading section, it is possible to rearrange the load mountain on the front side in the vehicle traveling direction in the loading section, and it is possible to shorten the movable distance of the load mountain toward the front side in the vehicle traveling direction. And thereby, even when there are few load mountains in the target loading section and the truck suddenly stops, it is possible to suppress excessive movement of each load mountain and suppress damage to the goods included in the load mountain.

[0056] The above (b-1) and (b-2) may be implemented individually or in combination with the above (a). Furthermore, it may be implemented in combination with the above (a) and (c) and / or (d) described later.

[0057] In the present invention, the load amount calculation target is identified based on the first loading image obtained by imaging the loading section from one side in the vehicle width direction. In addition to this, the load amount calculation target may be further identified based on the second loading image obtained by imaging the loading section from the other side in the vehicle width direction.

[0058] Hereinafter, as necessary, the load amount calculation target identified based on the first loading image may be referred to as the first load amount calculation target, and the load amount calculation target identified based on the second loading image may be referred to as the second load amount calculation target to distinguish the two. Also, as necessary, the camera for obtaining the first loading image may be referred to as the first camera, and the camera for obtaining the second loading image may be referred to as the second camera to distinguish the two. Furthermore, as necessary, the lower pallet of the load mountain in the second loading image may be referred to as the second lower pallet.

[0059] When identifying the second load quantity calculation target, specifically, the target loading part may be imaged from the other side opposite to the above-mentioned one side with respect to the target track by the second camera, and a second loading image may be acquired. Then, the loading part in the second loading image and the lower pallet of the load mountain in the second loading image may be compared. In addition, when a plurality of loading parts are shown in the second loading image, the largest loading part in the second loading image, that is, the loading part closest to the camera, may be set as the target loading part.

[0060] As described above, by comparing the target loading part and the second lower pallet, that is, the lower pallet of the load mountain in the second loading image, in the second loading image, it is possible to determine whether the second lower pallet is within a predetermined area of the target loading part. And when the second lower pallet is within the second predetermined area, it is possible to identify that the load mountain including the second lower pallet is the second load quantity calculation target.

[0061] In this case, the second predetermined area can be set in the same manner as the above-mentioned first predetermined area. As an example, in the second loading image of (c), a load mountain on the side closer to the reference, that is, on the side closer to the second camera, with the other side in the vehicle width direction as the reference can be cited as the second load quantity calculation target. According to (c), a load mountain on the side far from the reference with respect to the target loading part, that is, on the side far from the second camera, and further on one side in the vehicle width direction can be excluded from the second load quantity calculation target.

[0062] In addition, when identifying the load quantity calculation target according to the above (c), it is more preferable to use the second loading image obtained by imaging the loading part from an oblique direction including the other side than simply imaging the loading part from the other side of the target track. Specifically, it is the other side and the rear side, the other side and the front side, the other side and the rear side and the upper side, the other side and the front side and the upper side, etc.

[0063] In the above (c), the other end side in the vehicle width direction on the floor surface of the target loading part may be set as the second reference line, and the range where the distance from the second reference line is equal to or less than a predetermined value may be set as the second predetermined area.

[0064] For example, if the range where the distance from the above second reference line toward one side in the vehicle width direction is equal to or less than the width of the target loading part is set as the second predetermined area, the load mountains on one side in the vehicle width direction from the target loading part can be excluded, and the other load mountains can be set as the targets for calculating the second load quantity. Also, for example, if the range where the distance from the second reference line toward one side in the vehicle width direction is equal to or less than half of the width of the target loading part is set as the second predetermined area, only the load mountains on the other side in the vehicle width direction in the target loading part can be set as the targets for calculating the load quantity.

[0065] As another example, similar to the above (b), in (d) in the second loading image, an example can be given where the load mountains on the side closer to a predetermined position in the vehicle traveling direction are set as the target loading part. In the above (d), similar to the above (b), the load mountains on the side farther from the reference with respect to the target loading part, in other words, the load mountains on the front side or the rear side in the vehicle traveling direction from the target loading part, can be excluded from the targets for calculating the load quantity.

[0066] It is reasonable to perform the identification of the above second targets for calculating the load quantity after performing the identification of the first targets for calculating the load quantity, when it is determined that there are load mountains in the first loading image and there are no load mountains that are the targets for calculating the load quantity.

[0067] In addition, when performing the identification of the second targets for calculating the load quantity in addition to the identification of the first targets for calculating the load quantity, it is preferable that the range where the distance from the first reference line toward the other side in the vehicle width direction in the above (a) is smaller than the width of the target loading part is set as the predetermined area. And in this case, for example, it is more preferable that the range where the distance from the first reference line toward the other side in the vehicle width direction is equal to or less than half of the width of the target loading part is set as the predetermined area. This is because in the above (a), the load mountains identified as not being the first targets for calculating the load quantity can be identified as the second targets for calculating the load quantity in (c), and the identification of the targets for calculating the load quantity can be performed with high accuracy.

[0068] When identifying a second load calculation target in addition to the first load calculation target, it is necessary to suppress identifying the same load mountain as both the first and second load calculation targets. For this reason, it is preferable to set the first predetermined area small and the second predetermined area small. Specifically, it is preferable to set the distance Wa from the first reference line toward the other side in the vehicle width direction short in (a), and to set the distance Wc from the second reference line toward one side in the vehicle width direction short in (c).

[0069] More specifically, a range where Wa is 1 / 5 or less of the width of the target loading part is set as the first predetermined area and a range where Wc is 1 / 5 or less of the width of the target loading part is set as the second predetermined area, a range where Wa is 1 / 8 or less of the width of the target loading part is set as the first predetermined area and a range where Wc is 1 / 8 or less of the width of the target loading part is set as the second predetermined area, or a range where Wa is 1 / 10 or less of the width of the target loading part is set as the first predetermined area and a range where Wc is 1 / 10 or less of the width of the target loading part is set as the second predetermined area is particularly preferable.

[0070] Or, a range where Wa is 50 cm or less is set as the first predetermined area and a range where Wc is 50 cm or less is set as the second predetermined area, a range where Wa is 30 cm or less is set as the first predetermined area and a range where Wc is 30 cm or less is set as the second predetermined area, or a range where Wa is 20 cm or less is set as the first predetermined area and a range where Wc is 20 cm or less is set as the second predetermined area is particularly preferable.

[0071] Incidentally, the distance Wa and the distance Wc mentioned here may be calculated by a general method based on the loading image related thereto, that is, the first loading image or the second loading image.

[0072] Specifically, first, the dimensions or positional relationships of each part of the actual target loading section are measured in advance and obtained as these measured values. Also, the target loading section is imaged in advance by a first camera installed at a predetermined position to obtain a first loading image. Then, the measured values of the dimensions or positional relationships of each part in the actual target loading section are associated with the dimensions or positional relationships of each part in the first loading image. By doing so, it becomes possible to calculate the distances Wa and Wc based on the loading image. This also applies to the distance from the rear reference line toward the rear side in the vehicle traveling direction, the distance from the front reference line toward the front side in the vehicle traveling direction, and the distance from the front reference line toward the rear side in the vehicle traveling direction after the above description.

[0073] Note that as the dimensions, etc. of each part in the above-described first loading image, the measured values of the dimensions of each part in the first loading image may be used, or instead of the measured values, the image information of each part in the first loading image, for example, the number of pixels (specifically, the pixel count) of each part, etc. may be used.

[0074] In the present invention, by identifying each load mountain as described above and calculating the amount of the goods included in the load mountain identified as the first load amount calculation target, it becomes possible to accurately grasp the load amount of the target truck with high reliability. In addition to this, by calculating the amount of the goods included in the load mountain identified as the second load amount calculation target, it becomes possible to more accurately grasp the load amount of the target truck with higher reliability.

[0075] In the present invention, the step of identifying that the load mountain is the first load amount calculation target based on the first loading image, and the step of calculating the amount of the goods included in the load mountain identified as the first load amount calculation target can be performed by known load amount management elements. The same also applies to the step of identifying that the load mountain is the second load amount calculation target based on the second loading image, and the step of calculating the amount of the goods included in the load mountain identified as the second load amount calculation target.

[0076] The load management element only needs to have a function capable of performing the above-described process, and its structure and configuration are not particularly limited. Examples of the load management element include a computer including an arithmetic and control device such as a CPU and a storage device such as a memory or a hard disk.

[0077] Incidentally, a general loading section generally has a substantially rectangular shape when viewed from the side from one side or the other side in the width direction. In the first loading image acquired by the first camera and the second loading image acquired by the second camera, the loading section often has a distorted rectangular shape. This is because it is difficult to install the first camera and the second camera directly beside the loading section at the loading and unloading base, and cameras with a large angle of view capable of capturing the loading section as a rectangle in the first loading image and the second loading image are expensive.

[0078] In the present invention, in order to more reliably identify the first load calculation target and the second load calculation target based on the loading image, it is preferable to perform image processing to return the first loading image and the second loading image from a distorted rectangular shape to a rectangle or a parallelogram, and calculate the above-described distance Wa and distance Wc based on the first loading image and the second loading image after the image processing. The image processing can be performed by a known method, and as the program, application, or software used for the image processing, those having functions necessary for the target image processing may be appropriately used.

[0079] Hereinafter, the load calculation method, the luggage identification method, and the load calculation device of the present invention will be described with specific examples.

[0080] (Example 1) FIG. 1 is an explanatory diagram schematically explaining the positional relationship among the loading / unloading base, the truck, the loading section, and the load mountain in the load calculation method, the luggage identification method, and the load calculation device according to the first embodiment. FIG. 2 is an explanatory diagram schematically showing how the loading section looks when viewed from one side in the load calculation method, the luggage identification method, and the load calculation device according to the first embodiment. FIG. 3 is an explanatory diagram schematically showing a first loading image in the load calculation method, the luggage identification method, and the load calculation device according to the first embodiment. FIG. 4 is a flowchart for explaining the load calculation method and the luggage identification method according to the first embodiment. Hereinafter, each direction mentioned in the first embodiment shall mean each direction shown in each figure.

[0081] The load calculation device 1 according to the first embodiment includes two cameras 2 and a load management element 3 connected to the cameras 2.

[0082] Among these, the cameras 2 are installed outside the vehicle of the target truck 92t and on one side and the other side of the loading section 94, respectively, at the loading / unloading base 91 of the luggage 90.

[0083] The load management element 3 is a computer equipped with AI and is wirelessly connected to the cameras 2. The cameras 2 are controlled by the load management element 3 to capture a loading image as a video. The loading image acquired by the cameras 2 is transmitted to and stored in the load management element 3.

[0084] A large number of loading / unloading areas 93 for the trucks 92 are provided at the loading / unloading base 91. As shown in FIG. 1, each loading / unloading area 93 is a section where the truck 92 can park and is surrounded by lines drawn on the floor of the building. The truck 92 is inside the loading / unloading area 93 when entering the warehouse and exits the loading / unloading area 93 when leaving the warehouse.

[0085] Two cameras 2 are arranged for each loading / unloading area 93. The first camera 21, which is one of the cameras 2, is arranged on one side (left side), rear side, and upper side of the loading / unloading area 93, and the second camera 22, which is the other camera 2, is arranged on the other side (right side), rear side, and upper side of the loading / unloading area 93.

[0086] The first camera 21 acquires a first loading image obtained by imaging the loading section 94 of the target truck 92t from one side (the passenger seat side in the first embodiment) outside the vehicle. The second camera 22 acquires a second loading image obtained by imaging the loading section 94 of the target truck 92t from the other side (the driver's seat side in the first embodiment) outside the vehicle.

[0087] Here, a large number of loading areas 93 are provided in the loading and unloading base 91, and adjacent loading areas 93 are located relatively close to each other. Therefore, other trucks 92o may be parked in the vicinity of the target truck 92t parked in the loading area 93. In FIG. 1, the other truck 92o is adjacent to the other side of the target truck 92t.

[0088] In such an environment, when looking at the loading section 94 of the target truck 92t from one side in the vehicle width direction outside the vehicle, as shown in FIG. 2, other trucks 92o overlap on the other side in the vehicle width direction with respect to the target truck 92t, and the goods 90 and pallets 95 loaded on the other trucks 92o are reflected in the state of a cargo mountain 96 on the loading section 94 of the target truck 92t. Therefore, it is necessary to identify the actual cargo mountain 96 loaded on the target truck 92t as the CT for which the cargo quantity is to be calculated, and exclude the cargo mountain 96 derived from other trucks 92o as O outside the cargo quantity calculation target.

[0089] In the cargo quantity calculation method, the goods identification method, and the cargo quantity calculation device 1 of the first embodiment, the first camera 21 images the loading section 94 of the target truck 92t from one side, the rear side, and the upper side in the vehicle width direction outside the vehicle, and obtains the first loading image shown in FIG. 3. The obtained first loading image is transmitted to the cargo quantity management element 3.

[0090] The cargo quantity management element 3 obtains the outer shape 94OL of the loading section 94 in the first loading image. Then, by comparing the outer shape 94OL of the loading section 94 with the outer shapes of other loading sections 94, etc., the loading section 94 of the target truck 92t, that is, the target loading section 94t, is determined.

[0091] Also, the load management element 3 sets one side edge on the floor surface 94f of the target loading section 94t as the first reference line FS in the first loading image. Then, a range where the distance from the first reference line FS is equal to or less than a predetermined value is set as the first predetermined area FA. In the load calculation method, luggage identification method, and load calculation device 1 of the first embodiment, an area where the distance from the first reference line FS toward the other side is 20 cm or less as the measured value in the actual target loading section 94t is set as the first predetermined area FA.

[0092] Furthermore, the load management element 3 compares the lower pallet 95L of each load stack 96 with the target loading section 94t in the first loading image. When the lowermost part of the lower pallet 95L and one side edge SE are within the first predetermined area FA of the target loading section 94t, the load stack 96 is identified as a target for the first load calculation. Then, based on this identification result, the load management element 3 calculates the quantity of the luggage 90 included in the load stack 96.

[0093] For example, among many load stacks 96 shown in FIG. 3, a load stack 96f whose side edge SE is within the first predetermined area FA of the target loading section 94t is identified as a target for the first load calculation. On the other hand, a load stack 96s whose side edge SE is outside the first predetermined area FA of the target loading section 94t is excluded from the targets for the first load calculation. Regarding the load stacks 96 originating from other trucks 92o, since their side edges SE are outside the first predetermined area FA, they are naturally excluded from the targets for the first load calculation.

[0094] Accordingly, according to the load calculation method, luggage identification method, and load calculation device 1 of the first embodiment, it is possible to accurately identify whether the luggage 90 shown in the first loading image is the luggage 90 loaded on the target loading section 94t or the luggage 90 loaded on the loading section 94 of a truck 92o that is not a target for load calculation. As a result, it is possible to accurately calculate the load of the target truck 92t.

[0095] Further, in the load calculation method, the luggage identification method, and the load calculation device 1 of the first embodiment, the second camera 22 further acquires a second loading image obtained by imaging the loading part 94 of the target truck 92t from the other side (the driver's seat side in the first embodiment).

[0096] Based on the second loading image, the load management element 3 determines the target loading part 94t in the same method as described above. Although not shown, the load management element 3 uses the other side edge on the floor surface 94f of the target loading part 94t in the second loading image as the second reference line. Then, a range where the distance from the second reference line is equal to or less than a predetermined value is defined as the second predetermined area. In the load calculation method, the luggage identification method, and the load calculation device 1 of the second embodiment, an area where the distance from the second reference line toward one side is 20 cm or less as the measured value in the actual target loading part 94t is defined as the second predetermined area.

[0097] Furthermore, the load management element 3 compares the lower pallet 95L of each load stack 96 with the target loading part 94t in the second loading image. When the lowermost part of the lower pallet 95L and the other side edge are within the second predetermined area of the target loading part 94t, it is identified that the load stack 96 is the second load calculation target. Then, based on this identification result, the load management element 3 calculates the quantity of the luggage 90 included in the load stack 96.

[0098] For example, among many load stacks 96 shown in FIG. 3, for the load stack 96f, since the lowermost part of its lower pallet 95L and the other side edge are outside the second predetermined area of the target loading part 94t, it is excluded from the second load calculation target. On the other hand, for the load stack 96s, since the lowermost part of its lower pallet 95L and the other side edge are within the second predetermined area of the target loading part 94t, it is identified as the second load calculation target. Regarding the load stacks 96 originating from the other trucks 92o shown in FIG. 1, since they are not captured in the second loading image, they are naturally excluded from the second load calculation target.

[0099] According to the load calculation method, the luggage identification method, and the load calculation device 1 of the first embodiment, a load mountain 96 on one side of the target loading section 94t is identified as the first load calculation target, and the other load mountains 96 are excluded from the first load calculation target. Then, a load mountain 96 on the other side of the target loading section 94t is identified as the second load calculation target, and the other load mountains 96 are excluded from the second load calculation target.

[0100] Accordingly, according to the load calculation method, the luggage identification method, and the load calculation device 1 of the first embodiment, it is possible to accurately identify whether the luggage 90 shown in the first loading image is the luggage 90 loaded on the target loading section 94t or the luggage 90 loaded on the loading section 94 of the truck 92o that is not a load calculation target. In addition, it is possible to accurately identify whether the luggage 90 shown in the second loading image is the luggage 90 loaded on the target loading section 94t or the luggage 90 loaded on the loading section 94 of the truck 92 that is not a load calculation target.

[0101] Accordingly, according to the load calculation method, the luggage identification method, and the load calculation device 1 of the first embodiment, it is possible to calculate the load of the target truck 92t with very high accuracy.

[0102] Hereinafter, the load calculation method and the luggage identification method of the first embodiment will be described more specifically with reference to the flowchart shown in FIG. 4.

[0103] First, when the first loading image obtained by the first camera 21 is transmitted to the load management element 3, the load management element 3 starts the overall process (S1).

[0104] In S1, the load management element 3 extracts an image similar to the truck 92 from the first loading image and determines whether the image is the truck 92 (S2).

[0105] Next, load calculation (A) is performed on the target loading section 94t of the target truck 92t (S3). In S3, after the start of the process of load amount calculation (A) (SA1), first, learning for reading the first loading image is performed (SA2). Next, inference is performed based on the learning content of SA2 (SA3). Note that if a learned model in which learning corresponding to SA2 has been performed in advance is applied to the load amount management element 3 prior to the overall process S1 of this specification, SA2 may be skipped.

[0106] In SA3, specifically, the outer shape of the loading part 94 is obtained based on the coordinates of the four corners of the loading part 94 in the first loading image. If there is only one truck 92 in the first loading image, the truck 92 is determined as the target truck 92t. If there are a plurality of trucks 92 in the first loading image, the outer shapes of the loading parts 94 in each truck 92 are compared, the truck 92 with the largest loading part 94 is determined as the target truck 92t, and the loading part 94 of the target truck 92 is determined as the target loading part 94t. Further, the pallet 95 in the target loading part 94 is extracted.

[0107] After the above SA3, lower pallet determination (B) is performed on the pallet 95 in the target loading part 94 (SA4).

[0108] In SA4, specifically, after the start of the process of lower pallet determination (B) (SB1), first, the first reference line FS in the first loading image is detected (SB2). Next, the pallet 95 in the first loading image is extracted (SB3).

[0109] For one of the pallets 95 extracted in SB3, the outer shape of the pallet 95 is obtained based on the coordinates of its four corners. Then, the recognition frame coordinates (1) of the pallet 95, more specifically, the center coordinates in the vertical direction (i.e., the height direction) of the pallet 95 are obtained (SB4). Here, the four corners of the pallet 95 mean the four corners on the side surface of the pallet 95 shown in the first loading image taken from one side (the passenger seat side in Example 1) of the loading part 94. By obtaining the outer shape of the pallet 95 based on the four corners, the center coordinates in the height direction on the side surface of the pallet 95 can be obtained.

[0110] Next, the center coordinates (2) of the first reference line FS, more specifically, the vertical center coordinates on the first reference line FS are obtained (SB5).

[0111] Then, the recognition axis coordinates (1) of the pallet 95 obtained in SB4 are compared with the center coordinates (2) of the first reference line FS obtained in SB5, and it is determined whether (1) - (2)>0 and whether the (1) - (2) is less than or equal to the threshold value (SB6). As described above, in the first embodiment, the region where the distance from the first reference line FS to the other side is 20 cm or less in the actual measurement value in the actual target loading section 94t is defined as the first predetermined region FA. Therefore, the threshold value here is 20 cm in the actual measurement value.

[0112] When it is determined in SB6 that "(1) - (2)>0 and less than or equal to the threshold value" (Y in SB6), the pallet 95 is determined as the lower pallet 95L in the load mountain 96 which is the first load calculation target (SB7). Then, it is determined whether the determination SB7 of the lower pallet 95L has been completed for all the pallets 95 (SB8). If SB7 has been completed for all the pallets 95, the lower pallet determination (B) is ended (SB9). Also, if SB7 has not been completed for all the pallets 95, the process returns to SB3.

[0113] When it is not determined in SB6 that "(1) - (2)>0 and less than or equal to the threshold value" (N in SB6), the process proceeds to SB8. If SB7 has been completed for all the pallets 95, the lower pallet determination (B) is ended (SB9). Also, if SB7 has not been completed for all the pallets 95, the process returns to SB3.

[0114] After the completion of the process of the lower pallet determination (B) (SB9), the load mountain 96 including the lower pallet 95L is detected (SA5). As described above, since the load mountain 96 is the first load calculation target, the load included in the load mountain 96 is then calculated (SA6). After the completion of SA6, the load calculation (A) process is ended (SA7), and the result is registered in the database (S4), and the overall process is ended (S5).

[0115] In the load calculation method, the luggage identification method, and the load calculation device 1 of Example 1 of Example 1, further, for example, in the first loading image shown in FIG. 3, it is also possible to target the load mountain 96 on the side closer to the reference as the load to be calculated based on a predetermined position in the vehicle traveling direction.

[0116] For example, the front reference line FRS can be set as the front edge on the floor surface 94f of the target loading section 94t in the vehicle traveling direction. And the range where the distance from the front reference line FRS toward the rear side in the vehicle traveling direction is half or less of the length of the target loading section 94t in the vehicle traveling direction can also be set as the third predetermined area TA.

[0117] In this case, the load mountain 96 at the lowermost part of the lower pallet 95L and whose rear edge is within the third predetermined area TA can be identified as the third load calculation target, and other load mountains 96 can be excluded from the third load calculation target. Thereby, it is possible to accurately identify whether the luggage 90 shown in the first loading image is the luggage 90 loaded on the target loading section 94t or the luggage 90 loaded on the loading section 94 of the truck 92 that is not the load calculation target, and it is possible to accurately calculate the load of the target truck 92t.

[0118] In addition, in this case, when there are few or no load mountains 96 that are the third load calculation target and there are load mountains 96 that are not the third load calculation target, it is possible to alert the operator or give an instruction to the work robot to rearrange the load mountain 96. Thereby, when there is a load mountain 96 that is not the third load calculation target in the target loading section 94t, the load mountain 96 can be rearranged to the front side in the vehicle traveling direction in the target loading section 94t. Thereby, the movable distance of the load mountain 96 toward the front side in the vehicle traveling direction can be shortened, and as a result, even when there are few load mountains 96 in the target loading section 94t and the truck 92 makes an emergency stop, excessive movement of each load mountain 96 can be suppressed, and damage to the luggage 90 included in the load mountain 96 can be suppressed.

[0119] Although the present invention has been described above, the present invention is not limited to the above-described embodiments and the like, and it is possible to appropriately extract and combine the elements described in the above embodiments and the like and to make various changes without departing from the gist of the present invention. In addition, the specification of the present invention discloses a technical idea in which the matters described in each claim are appropriately combined, not limited to the citation relationship of each claim at the time of filing the application.

Explanation of Signs

[0120] 1: Truck load calculation device 2: Camera 3: Load management element 90: Freight 92: Truck 94: Loading section 94f: Floor surface of the loading section 95: Pallet 95L: Pallet at the lowermost stage of the cargo stack 96: Cargo stack FS: First reference line

Claims

1. A method for calculating the load quantity in the loading section of a truck, comprising: acquiring a first loading image obtained by imaging the loading section from the outside of the vehicle and on one side in the vehicle width direction with respect to the truck; comparing, in the first loading image, the loading section with the lowermost pallet in the load mountain on which the goods are loaded; when the pallet is within a predetermined area of the loading section, identifying that the load mountain is an object for load quantity calculation, and calculating the quantity of the goods included in the load mountain. A method for calculating the load quantity of a truck.

2. Using the edge on one side of the floor surface of the loading section as a first reference line, setting a range where the distance from the first reference line is a predetermined value or less as the predetermined area, when the lowermost part of the pallet in the load mountain and the edge on one side are within the predetermined area, identifying that the load mountain is an object for load quantity calculation. The method for calculating the load quantity of a truck according to Claim 1.

3. Furthermore, acquiring a second loading image obtained by imaging the loading section from the other side opposite to the one side with respect to the truck; comparing, in the second loading image, the loading section with the lowermost pallet in the load mountain; when the pallet is within a second predetermined area of the loading section, identifying that the load mountain is an object for load quantity calculation, and calculating the quantity of the goods included in the load mountain. The method for calculating the load quantity of a truck according to Claim 1 or Claim 2.

4. A method for identifying whether the goods loaded on the loading section of a truck are an object for load quantity calculation, comprising: acquiring a first loading image obtained by imaging the loading section from the outside of the vehicle and on one side in the vehicle width direction with respect to the truck; comparing, in the first loading image, the loading section with the lowermost pallet in the load mountain on which the goods are loaded; when the pallet is within a predetermined area of the loading section, identifying that the load mountain is an object for load quantity calculation. A method for identifying goods on a truck.

5. A load quantity calculation device for calculating the load quantity in the loading section of a truck, comprising: a camera installed on the outside of the vehicle and on one side in the vehicle width direction with respect to the truck, imaging the loading section to acquire a first loading image; a load management element that compares, in the first loading image, the loading section with the lowermost pallet in the load mountain on which the goods are loaded, and when the pallet is within a predetermined area of the loading section, identifies that the load mountain is an object for load quantity calculation and calculates the quantity of the goods included in the load mountain. A load quantity calculation device for a truck.

Images