Object identification device and object identification method

Abstract

To provide an object identification device that can efficiently and accurately identify objects used for product transport. [Solution] The object identification device 10 of the present invention identifies the type of object, such as a bobbin 3, using images of multiple bobbin 3, etc., contained in a packaging box 2, in image information containing images of multiple bobbin 3, etc., using pre-learned images of bobbin 3, etc., measures the distance of the identified multiple bobbin 3, etc., from the object identification device 10, and compares the height of each bobbin 3, etc., calculated using the distance measurement results of each bobbin 3, etc., with the height of a bobbin 3, etc., that has been registered in advance, thereby correcting the identification result of the type of object, such as a bobbin 3, etc., identified using images of bobbin 3, etc.

Description

Technical Field

[0001] The present invention relates to an object identification device and an object identification method for identifying an object housed in a packaging box.

Background Art

[0002] When shipping and transporting goods from a warehouse or factory, transport objects such as pallets and racks are used. After the goods are transported to the destination, the transport objects such as pallets and racks are returned to the shipping origin and reused. In order to improve the recovery rate of transport objects such as pallets and racks, technologies for confirming the location of pallets and the like have also been proposed (see, for example, Patent Document 1).

[0003] As objects for transporting goods, various forms are used according to the goods to be transported. For example, in the textile industry, bobbins are used as objects for transporting fibers from production processes such as textile factories to textile sales destinations such as clothing manufacturers. Bobbins used for transportation to textile sales destinations are required to be returned to the sales origin of the transported fibers. Since bobbins are not inexpensive, a deposit may be imposed on sales destinations such as clothing manufacturers.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Here, when verifying the return of used transport objects such as bobbins at the point of sale of textile products, it is necessary to identify a large number of transport objects such as bobbins, which come in various types depending on the supplier and the type of textile product, and there is a need to streamline the identification process of transport objects such as bobbins. Recently, technology has been put into practical use that uses cameras installed in designated locations to acquire images of transport objects such as bobbins and identify them, but since many transport objects such as bobbins have similar shapes, the identification process is often ultimately carried out by visual inspection by workers.

[0006] This invention was made to solve the above-mentioned problems and aims to provide an object identification device that can efficiently and accurately identify objects for the transport of goods. [Means for solving the problem]

[0007] To solve the problems described above, the object identification device of the present invention comprises: an imaging unit that acquires image information including images of a plurality of objects contained in a packaging box; an object identification unit that identifies the type of object based on pre-learned images of the object in the image information including images of the plurality of objects acquired by the imaging unit; and a distance measuring unit that measures the distance from the imaging unit to each of the plurality of objects identified by the object identification unit. The object identification unit corrects the type of object identified based on the image of the object by comparing the height of each of the plurality of objects calculated using the measurement results of the distance measuring unit with the height of each of the plurality of objects that have been registered in advance.

[0008] To solve the problems described above, the present invention provides an object identification method that is performed in an object identification device comprising: an imaging unit that acquires image information containing images of a plurality of objects; an object identification unit that identifies the type of the object; and a distance measuring unit that measures the distance of each of the plurality of objects from the imaging unit, the method comprising: acquiring image information containing images of a plurality of objects contained in a packaging box; the object identification unit identifying the type of the object based on pre-learned images of the object in the image information containing images of the plurality of objects acquired by the imaging unit; and the distance measuring unit measuring the distance of each of the plurality of objects identified by the object identification unit from the imaging unit, wherein the object identification unit corrects the type of the object identified based on the images of the object by comparing the height of each of the plurality of objects calculated using the measurement results of the distance measuring unit with the height of each of the plurality of objects registered in advance. [Effects of the Invention]

[0009] According to the present invention, it is possible to provide an object identification device that can efficiently and accurately identify objects for the transportation of goods. [Brief explanation of the drawing]

[0010] [Figure 1] Figure 1 shows an example of a system configuration equipped with an object identification device according to an embodiment of the present invention. [Figure 2] Figure 2 shows an example of a functional block of an object identification device according to an embodiment of the present invention. [Figure 3] Figure 3 is a diagram illustrating object identification using an image of an object in an embodiment of the present invention. [Figure 4] Figure 4 is a diagram illustrating the modification of object identification results using the distance measurement results of objects in an embodiment of the present invention. [Figure 5] Figure 5 shows an example of the operation flow of the object identification method according to an embodiment of the present invention. [Modes for carrying out the invention]

[0011] Embodiments of the present invention will be described below with reference to the drawings. It should be noted that the present invention can be implemented in various embodiments and is not limited to the embodiments described below. In particular, the following embodiments describe the use of bobbins for transporting fibers as an example of an object to which the present invention applies for transporting goods, but the application of the present invention is not limited to bobbins for transporting fibers. The present invention can be applied to various transport objects other than bobbins for transporting fibers. It is particularly useful when transporting goods using multiple articles of different heights as transport objects, and when performing verification work to return the transport objects at the destination of the goods.

[0012] <Configuration of the object recognition system> Various forms of objects are used to transport goods, depending on the goods being transported. For example, in the textile industry, bobbins are used to transport fibers from production processes such as textile factories to sales destinations such as clothing manufacturers. Bobbins used for transporting fibers to sales destinations are required to be returned to the supplier of the transported fibers. At the fiber sales destination, when verifying the return of used bobbins, a technology has been put into practical use in which images of the bobbins are taken with cameras installed in designated locations, and the captured image information is used to identify the bobbins. In such verification work for transported objects, images of multiple bobbins in packaging boxes such as cardboard boxes are taken to improve the efficiency of the verification process.

[0013] Figure 1 shows an example of a system configuration equipped with an object recognition device according to an embodiment of the present invention. The object recognition system 1 in Figure 1 includes an object recognition device 10 that acquires image information containing images of multiple bobbins 3 housed in a packaging box 2 such as a cardboard box, and identifies the type of bobbin 3 based on pre-learned bobbin images in the acquired image information containing images of multiple bobbins 3.

[0014] In the system configuration shown in Figure 1, an object identification device 10 is installed in a predetermined location, and an image of the bobbin is acquired using the camera mounted on the object identification device 10. The acquired image information is then used to identify the bobbin. In this verification process, to improve the efficiency of the verification process, images of multiple bobbins contained in a packaging box 2 such as a cardboard box are acquired. As shown in the "top view" of Figure 1, the image information of the bobbin 3 viewed from above is used to identify the bobbin 3.

[0015] In the actual verification process for transporting bobbins and other transport objects, it is necessary to identify a large number of transport objects of various types, depending on the fiber supplier and the type of fiber. Since many transport objects such as bobbins have similar shapes, there is a risk of misidentifying the type of transport object, such as bobbin 3. In particular, when transport objects such as bobbin 3 are identified using image information of the transport object viewed from above, the likelihood of misidentifying the type of transport object increases.

[0016] The object identification device 10 of this embodiment is configured to correct the type of transport object identified based on images of transport objects such as bobbins, by measuring the distance from the object identification device to each of the multiple transport objects such as bobbins identified using image information, and comparing the height of each of the multiple transport objects such as bobbins calculated using the measurement results of each transport object with the height of each of the multiple transport objects such as bobbins that have been registered in advance.

[0017] According to this embodiment, by correcting the identification result of transport objects such as bobbins using image information with height information of transport objects such as bobbins, it is possible to provide an object identification device that can efficiently and accurately identify transport objects such as bobbins, even when identifying a large number of transport objects of various types.

[0018] <Configuration of Object Identification Device> FIG. 2 is an example of a functional block of an object identification device according to an embodiment of the present invention.

[0019] The object identification device 10 includes an imaging unit 11 that acquires image information including images of objects for transportation such as a plurality of bobbins 3 housed in a packaging box 2, a distance sensor unit 12 that detects the distance from the imaging unit 11 to the objects for transportation such as the bobbin 3, and uses the image information including the images of the objects for transportation such as the plurality of bobbins 3 acquired by the imaging unit 11 to identify the type of the objects for transportation such as the bobbin 3. A central processing unit that corrects the identification result of the type of the object for transportation such as the bobbin 3 using the image information, a storage unit 15 for storing various information such as the identification result of the object for transportation such as the bobbin 3, an input unit 16 for an operator to input information, and a display unit 17 for displaying the identification result of the object for transportation such as the bobbin 3.

[0020] As the imaging unit 11, a general-purpose RGB camera can be used. As the distance sensor unit 12, a general-purpose distance measurement sensor unit such as a TOF (Time of Flight) sensor or a LiDAR sensor can be used. As the object identification device 10 including the imaging unit 11 and the distance sensor unit 12, for example, a digital camera equipped with a CPU, a memory, an RGB camera, and a TOF (Time of Flight) sensor can be used.

[0021] The central processing unit functions as a distance measurement unit that measures the distance to the objects for transportation such as the bobbin 3 using the distance sensor unit 12, and in the image information including the images of the plurality of bobbins 3 acquired by the imaging unit, identifies the type of the bobbin 3 based on the pre-learned images of the bobbin, and uses the distance information to the objects for transportation such as the bobbin 3 measured by the distance measurement unit 13 to correct the identification result of the type of the objects for transportation such as the bobbin 3 using the image information. It functions as an object identification unit 14.

[0022] The distance measuring unit 13 measures the distance to a transport object such as a bobbin 3 using the distance sensor unit 12. The distance measuring unit 13 measures the distance from the imaging unit 11 to each of the multiple transport objects such as bobbins 3 identified by the object identification unit 14. The distance measurement results for the multiple transport objects such as bobbins 3 are used when calculating the height of each of the multiple transport objects such as bobbins 3.

[0023] The object identification unit 14 identifies the type of transport object, such as a bobbin 3, based on pre-learned images of transport objects, such as bobbins 3, in image information that includes images of multiple transport objects such as bobbins 3 acquired by the imaging unit 11. It also corrects the type of transport object identified based on the images of the transport objects by comparing the height of each of the multiple transport objects calculated using the measurement results of the distance measuring unit 13 with the height of each of the multiple transport objects such as bobbins that have been registered in advance.

[0024] The memory unit 15 stores pre-learned image information of transport objects such as bobbins 3, pre-registered height information of multiple transport objects such as bobbins, and the identification results of transport objects such as bobbins 3. The pre-learned image information of transport objects such as bobbins 3 is used to identify images of transport objects such as bobbins 3 in the image information acquired by the imaging unit 11. The pre-registered height information of multiple transport objects such as bobbins is used to correct the identification information of transport objects such as bobbins 3 by comparing it with the heights of each of the multiple transport objects calculated using the measurement results of the distance measuring unit 13.

[0025] <Object identification using image information> Figure 3 is a diagram illustrating object identification using an image of an object in an embodiment of the present invention.

[0026] Figure 3 shows an example of the results of identifying the type of transport object, such as the nine bobbins 3, stored in a packaging box 2 as illustrated in Figure 1, by acquiring image information. In the example in Figure 1, three "Bobbin #A", three "Bobbin #B", and three "Bobbin #C" are stored in the packaging box 2. The object identification device 10 identifies bobbins #A, #B, and #C by comparing the image information acquired by the imaging unit 11 with pre-learned image information of "Bobbin #A", "Bobbin #B", and "Bobbin #C".

[0027] Figure 3 shows the number of bobbins in each of the "Bobbin #A", "Bobbin #B", and "Bobbin #C" after identification. As a result of the identification, as shown in the table in Figure 3, 2 "Bobbin #A", 3 "Bobbin #B", and 4 "Bobbin #C" were identified. In the example in Figure 3, one of the "Bobbin #A" was mistakenly identified as "Bobbin #C" as a result of image-based identification.

[0028] In the table in Figure 3, "Bobbin Height" refers to the pre-registered height information for "Bobbin #A", "Bobbin #B", and "Bobbin #C". The object recognition device 10 uses the pre-registered height information for "Bobbin #A", "Bobbin #B", and "Bobbin #C" to correct the image information-based identification result of bobbin 3.

[0029] <Correction of object type using distance measurement results> Figure 4 is a diagram illustrating the modification of bobbin identification using distance measurement results for transport objects such as bobbins in an embodiment of the present invention. Figure 4 is an example of the results of measuring the distance between "Bobbin #A", "Bobbin #B", and "Bobbin #C" in a packaging box 2 containing the three types "Bobbin #A", "Bobbin #B", and "Bobbin #C" illustrated in Figure 1.

[0030] The object recognition device 10 uses a distance sensor unit 12 to measure the distance between the imaging unit 11 and "bobbin #A", "bobbin #B", and "bobbin #C". In the example in Figure 4, "distance A", "distance B", and "distance C" are measured as the distances between the imaging unit 11 and "bobbin #A", "bobbin #B", and "bobbin #C", respectively.

[0031] "Distance Z" is the distance between the imaging unit 11 and the bottom of the packaging box 2. "Distance Z" is used when calculating the heights of "Bobbin #A", "Bobbin #B", and "Bobbin #C" using "Distance A", "Distance B", and "Distance C". In the example in Figure 4, the heights of Bobbin #A, Bobbin #B, and Bobbin #C can be calculated using "Distance Z - Distance A", "Distance Z - Distance B", and "Distance Z - Distance C", respectively.

[0032] Regarding "distance Z," when measuring the distance to "bobbin #A," "bobbin #B," and "bobbin #C," it may be measured from the position of the conveyor belt or similar device on which the packaging box 2 is placed when identifying transport objects such as bobbin 3, or it may be measured in advance based on the position where the object identification device 10 is installed and the position of the conveyor belt or similar device on which the packaging box 2 is placed when identifying transport objects such as bobbin 3.

[0033] In this embodiment, the height of each of the transport objects, such as bobbins, is calculated using the measurement results of the distance measuring unit 13, and the type of transport object, such as bobbins, identified based on an image of the transport object, is corrected by comparing the heights of each of the transport objects, such as bobbins, that have been registered in advance.

[0034] The image-based identification of transport objects such as bobbins is corrected by comparing the calculated "Distance Z - Distance A," "Distance Z - Distance B," and "Distance Z - Distance C," which are obtained using the measured distances of each transport object such as bobbins, with the pre-registered heights of "Bobbin #A," "Bobbin #B," and "Bobbin #C," which are "40cm~45cm," "25cm~30cm," and "50cm~55cm."

[0035] In Figures 3 and 4, a predetermined range of values ​​is pre-registered as height information for objects such as bobbins. This is to account for measurement errors in the distance sensor unit 12 and measurement errors due to the positional relationship between the object identification device 10 and the packaging box 2. If the height of each object, such as a bobbin, calculated using the distance information falls within the predetermined range registered in Figures 3 and 4, the image-based identification result of the bobbin 3 can be determined to be correct. On the other hand, if the height of each object, such as a bobbin, calculated using the distance information falls outside the predetermined range registered in Figures 3 and 4, the image-based identification result of the bobbin 3 can be determined to be incorrect.

[0036] In the image recognition results explained in Figure 3, one of the "Bobbin #A" is incorrectly identified as "Bobbin #C". As shown in Figure 4, by comparing the heights, it can be detected that the calculated height of one of the incorrectly identified Bobbin 3 differs from the pre-registered height, so the incorrectly identified "Bobbin #C" can be corrected to the correct "Bobbin #A".

[0037] <Operation Flow of Object Identification Device> The operation of the identification code detection method performed in the object identification device 10 will be explained using Figure 5. Figure 5 is an example of the operation flow of the object identification method according to an embodiment of the present invention.

[0038] Each step of the object identification method is performed in an object identification device 10 which includes an imaging unit that acquires image information containing images of multiple transport objects such as bobbins 3, a bobbin identification unit that identifies the type of transport object such as bobbins 3, and a distance measuring unit that measures the distance from the imaging unit to each of the multiple transport objects such as bobbins 3.

[0039] The imaging unit of the object identification device 10 acquires image information that includes images of multiple transport objects such as bobbins 3 contained in a packaging box (S1-1).

[0040] The object identification unit of the object identification device 10 identifies the type of transport object, such as a bobbin 3, based on pre-learned images of transport objects such as bobbins 3, in the image information which includes images of multiple transport objects such as bobbins 3 acquired by the imaging unit (S1-2).

[0041] The distance measuring unit of the object identification device 10 measures the distance from the imaging unit to each of the multiple transport objects such as bobbins 3 identified by the object identification unit (S1-3).

[0042] The object identification unit of the object identification device 10 calculates the height of each of the multiple transport objects such as bobbins using the measurement results from the distance measuring unit (S1-4), and determines whether the calculated height of the transport objects such as bobbins is within a predetermined range of the height information of the multiple transport objects such as bobbins 3 that have been registered in advance (S1-5).

[0043] If the calculated height of the transport object, such as a bobbin, falls within a predetermined range of the height information of each of the multiple transport objects, such as bobbins 3, that have been registered in advance (S1-5: YES), then it can be determined that the identification result of the transport object, such as a bobbin 3, using the image is correct (S1-7).

[0044] On the other hand, if the result calculated using distance information is outside the predetermined range (S1-5: NO), it is determined that the identification result of the transport object such as bobbin 3 using the image is incorrect, and the identification result of the type of transport object such as bobbin using the image of the transport object such as bobbin is corrected (S1-6).

[0045] The object identification device of this embodiment is configured to correct the type of transport object identified based on the image of the transport object, such as a bobbin, by measuring the distance from the object identification device to each of the transport objects identified using the image information, and comparing the height of each transport object calculated using the measurement results of each transport object with the height of each transport object registered in advance.

[0046] According to this embodiment, by correcting the identification result of transport objects such as bobbins using image information with height information of transport objects such as bobbins, it is possible to provide an object identification device that can efficiently and accurately identify transport objects such as bobbins of various types, even when identifying a large number of transport objects such as bobbins. [Explanation of Symbols]

[0047] 1...Object identification system, 2...Packaging box, 3...Bobbin, 10...Object identification device, 11...Imaging unit, 12...Distance sensor unit, 13...Distance measuring unit, 14...Object identification unit, 15...Storage unit, 16...Input unit, 17...Display unit.

Claims

1. An imaging unit that acquires image information including images of multiple objects contained in a packaging box, In the image information containing images of the multiple objects acquired by the imaging unit, an object identification unit identifies the type of the object by comparing it with images of the objects that have been registered in advance. For each of the plurality of objects identified by the object identification unit, a distance measuring unit measures the distance from the imaging unit, Equipped with, The object identification unit is By comparing the height of each of the multiple objects calculated using the measurement results of the distance measuring unit with the height of each of the multiple objects registered in advance, the type of the object identified based on the image of the object is corrected. Object identification device.

2. An object identification method performed in an object identification device comprising: an imaging unit that acquires image information containing images of multiple objects; an object identification unit that identifies the type of the object; and a distance measuring unit that measures the distance of each of the multiple objects from the imaging unit, the method being performed The steps include obtaining image information that includes images of multiple objects contained in a packaging box, The object identification unit identifies the type of object by comparing the image information, which includes images of the plurality of objects acquired by the imaging unit, with images of the objects that have been registered in advance. The distance measuring unit measures the distance from the imaging unit to each of the plurality of objects identified by the object identification unit. It has, The object identification unit is By comparing the height of each of the multiple objects calculated using the measurement results of the distance measuring unit with the height of each of the multiple objects registered in advance, the type of the object identified based on the image of the object is corrected. Object identification method.

Images