Method and apparatus for detecting basic data of an object

Abstract

This application relates to a method and apparatus for detecting basic data of an object. The apparatus includes: a balance; a camera for recording images of the object; an evaluation unit having a volume determination unit for determining the object's volume; a starter unit that automatically determines a start time point for recording basic data related to the object's volume when the object is placed on the balance; a memory for storing the recorded basic data; and an output unit for outputting the basic data. The starter unit is designed to determine the start time point based on the images recorded by the camera, and for this purpose, the starter unit has an image evaluation unit designed to perform object classification, through which foreign objects, in addition to the object itself, are also identified as permitted or disallowed foreign objects.

Description

Technical Field

[0001] This invention relates to a method and apparatus for detecting the basic data (Stammdaten) of an object. Background Technology

[0002] The basic data of an object mainly includes its volume and weight. To detect these data, according to existing technology, the object is placed on a balance and measured using a volume recording system. For accurate basic data detection, the object must be freely positioned on the balance, with no foreign objects in the surrounding environment, and then the measurement is performed automatically using an optical volume detection system. If a foreign object is present in the camera's field of view, such as the hand or arm of a person placing the object on the balance, the volume detection system will no longer be able to detect the object correctly. This also applies to the weighing process; during weighing, the object on the balance should not be touched to ensure that the weight is not tampered with.

[0003] From Metrilus GmbH, a basic data detection system called S110 / 120 is known. This system includes a balance for detecting weight, a 3D camera for determining volume, and a barcode reader. If the barcode reader identifies an object, and the volume detection system does not detect any foreign objects, the 3D camera identifies the object on the balance and initiates basic data detection.

[0004] Therefore, in order to measure an object, the object must be placed freely within the defined area. In this case, "freely" means without contact with other objects. Thus, during the measurement process, it is usually necessary to ensure that the measurement area is completely free of external objects. External objects here also include, for example, hands and other objects that extend into the measurement area from the outside without touching the object.

[0005] The disadvantages of doing this are increased workload during the measurement process, as well as the resulting speed disadvantage or measurement errors due to measuring foreign objects. Summary of the Invention

[0006] Therefore, the object of the present invention is to provide improved methods and new devices for detecting basic data of an object, and in particular, to use them to detect basic data more quickly and reliably.

[0007] This objective is achieved by a method having the features described below and a device having the features described below.

[0008] The method for detecting basic data of an object according to the present invention includes the following steps:

[0009] Place the object on the scale.

[0010] The following steps are used to determine the start time point for recording basic data:

[0011] Use a camera to record an image of the object with depth values.

[0012] Continuous regions in the image are measured, and these continuous regions then form the first temporary object.

[0013] The first temporary objects are classified into valid and invalid classes. Specifically, all first temporary objects located inside the image and not in contact with the image edges are classified as valid, while those in contact with the edges are classified as invalid.

[0014] By evaluating depth values ​​and forming new sub-regions with depth values ​​exceeding a depth threshold, the first temporary object in the invalid class is analyzed.

[0015] Examine these sub-regions to find contiguous regions that form a second temporary object.

[0016] If the second temporary object does not touch the image edge, then move the second temporary object to the valid class.

[0017] Find the overlap (Überdeckung) between temporary objects in valid classes and temporary objects remaining in invalid classes, where the overlap must exceed an overlap threshold.

[0018] If there is overlap, the relevant temporary object is moved from the valid class to the invalid class.

[0019] Define the temporary object in the valid class as the object to be measured.

[0020] If there are no objects in the valid class, output an error signal; otherwise...

[0021] After the startup time point determined in this way, the volume of the startup object is detected.

[0022] Record basic data.

[0023] Store the basic data that is recorded.

[0024] Output basic data.

[0025] By classifying foreign objects into allowed and disallowed foreign objects, data acquisition can now be performed even when the foreign object (i.e., allowed foreign object) is within the camera's field of view. This increases throughput (Durchsatz) because data detection begins when the hand is released from the object placed on the scale but is still within the camera's field of view.

[0026] By classifying foreign objects, data can be detected more reliably. This also improves the analysis of potential errors.

[0027] An advantageous design of the method according to the invention involves recording and evaluating a depth value histogram while analyzing a first temporary object to evaluate its depth value and forming a new sub-region with a depth value exceeding a depth value threshold. This facilitates the classification of foreign objects into permitted and disallowed foreign objects.

[0028] Because 3D cameras often encounter areas in the image where depth values ​​are missing, making meaningful evaluation impossible, a further extension of this invention includes searching for overlaps between temporary objects in the valid class and areas lacking depth values. If overlap is determined, the relevant temporary objects are moved from the valid class to the invalid class.

[0029] In a further extension of the invention, it is meaningful to output an error signal if there are no objects in the valid class at the end of the evaluation, which means that basic data detection is not possible.

[0030] This objective is also achieved by a device according to the invention for detecting basic data of an object, the device having the following characteristics:

[0031] Balance,

[0032] A camera, used to record images of objects.

[0033] The evaluation unit includes a volume determination unit for determining the volume of the object.

[0034] The initiator unit automatically determines the start-up time for recording basic data related to the object's volume when the object is placed on the balance.

[0035] A memory, which is used to store the basic data that has been recorded.

[0036] The output unit is used to output basic data.

[0037] in:

[0038] The initiator unit is designed to determine the initiation time point based on the images recorded by the camera, and for this purpose, the initiator unit has an image evaluation unit, which is designed to perform object classification, by which foreign objects are identified as either allowed or disallowed foreign objects in addition to the objects themselves.

[0039] In the first extended scheme, the initiator unit is designed to output the start-up time point as early as possible, even while permitted foreign objects are still within the image region. This further improves throughput.

[0040] To facilitate classification, the camera generates depth values, i.e., records 3D images, so that the image evaluation unit can consider the depth values ​​when classifying objects, which is beneficial for classification.

[0041] In an extended embodiment of the invention, the evaluation unit is designed to output a message, i.e., an error signal, if the start-up time cannot be determined within the waiting time. In this way, the system user can quickly identify that basic data has not yet been detected. Attached Figure Description

[0042] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. In the drawings:

[0043] Figure 1 A schematic diagram of the device according to the present invention is shown;

[0044] Figures 2 to 6 A schematic top view of the camera's field of view is shown, with foreign objects in different positions.

[0045] Figure 7 A schematic top view of the camera's field of view with image defects is shown;

[0046] Figure 8 A flowchart of the method according to the present invention is shown. Detailed Implementation

[0047] The device 10 according to the invention is used to detect basic data of an object 12. In addition to the mechanical structural element 13, the device 10 also includes a balance 14, a camera 16 for recording images of the object 12, and an evaluation unit 18. The evaluation unit 18 has a volume determination unit 20 for determining the volume of the object 12. The weight of the object 12 is detected using the balance 14, and the volume is detected using the volume determination unit 20. The data thus determined can ultimately be output to the output terminal 24 via the output unit 22 as basic data, or at least as part of the basic data. A memory 28 is provided for storing or caching the recorded basic data.

[0048] The evaluation unit 18 also includes a starter unit 26, which automatically determines the start time point for recording and measuring basic data related to the object's volume when the object 12 is placed in the field of view 32 of the camera 16. For this purpose, the starter unit 26 is designed to determine the start time point based on the image recorded by the camera 16, and the starter unit 26 has an image evaluation unit 30, which is designed to perform object classification, classifying foreign objects 50, in addition to object 12, as either permitted or prohibited foreign objects.

[0049] The starter unit 26 is designed to output the start time point as early as possible, and can output the start time point while the foreign object 50 classified as allowed is still in the image area 32 of the camera 16.

[0050] Camera 16 is designed as a 3D camera, which, in addition to the 2D image with image edges 33, also measures the depth value of each pixel, thereby recording a 3D image as a whole.

[0051] exist Figure 8 A flowchart illustrating at least a portion of an embodiment of the method 100 according to the present invention is shown, and reference is also made below to... Figures 3 to 7 To elaborate. Figures 3 to 7 Top views of the balance 14 and the object 12 placed on it, as seen from the perspective of camera 16, are shown. In these figures, the edge 33 of the field of view 32 is shown as a dashed line, which is the image edge 33 in the 2D image.

[0052] Method 100 mostly runs in the initiator unit 26, with the purpose of initiating the detection of the volume of object 12 by means of a start signal at the start time, thereby initiating the basic data detection.

[0053] In the first step 110, the object is placed on the balance 14.

[0054] Then, the start-up time point is determined, at which a start-up signal is generated that records the basic start-up data. The determination of the start-up time point is accomplished through the following steps.

[0055] In step 120, the camera 16 records an image of the object 12 with depth values.

[0056] In step 130, based on depth values, continuous regions in the image are determined. These regions, in the context of this patent application, can be individual pixels or groups of pixels in the image. These continuous regions form a first temporary object. For example, in... Figure 3 In this context, the continuous region is object 12. Figure 6 In the image evaluation, the continuous region will be the hand with an arm extending into the field of view 32 (i.e., the foreign object 50) along with the object 12 partially overlapped by the hand. The entire image evaluation is performed in the image evaluation unit 30.

[0057] In the next step 140, the first temporary objects are classified into a valid class and an invalid class. All first temporary objects located within the image and not in contact with the image edge 33 are classified as valid, while first temporary objects in contact with the edge are classified as invalid.

[0058] For example, in this way:

[0059] According to Figure 3In this case, object 12 is classified as a valid class;

[0060] According to Figure 4 In this case, object 12 is classified as a valid class, while foreign object 50 is classified as an invalid class;

[0061] According to Figure 5 and Figure 6 In this case, there is only one first temporary object, because the external object 50 comes into contact with at least the object 12 to be measured, and the two then form the first temporary object. The first temporary object is classified as an invalid class.

[0062] exist Figure 7 The diagram illustrates a case where no foreign object 50 exists, but a region 52 without a depth value exists within the field of view 32. This particular error case will be considered later.

[0063] In step 150, the first temporary objects classified as invalid are analyzed. Here, the depth values ​​are evaluated to see if they exceed a depth value threshold. If so, these regions form new sub-regions. Preferably, in order to evaluate the depth values ​​and to form new sub-regions with depth values ​​exceeding the depth value threshold, a depth value histogram is recorded and evaluated in evaluation unit 18.

[0064] In the next step, the sub-regions are examined to determine if they are continuous; these continuous regions then form the second temporary object. Typically, peaks in the depth value histogram correspond to the second temporary object. Therefore, in this way, and using the depth value histogram, based on... Figure 5 and Figure 6 In this case, object 12 and foreign object 50 are separated into two second temporary objects.

[0065] In the next step 170, it is checked whether these second temporary objects are in contact with image edge 33. If the second temporary object is not in contact with image edge 33, it is moved from the invalid class to the valid class. For example, this would be Figure 5 Object 12 in the context.

[0066] Therefore, so far, temporary objects classified as valid class 180 or invalid class 190 have been found in the method according to the invention.

[0067] In the subsequent step 200, it is checked whether the temporary objects of the valid class overlap with the temporary objects remaining in the invalid class. Here, it is meaningful to consider the overlap as overlapping if the overlapping area is large enough, i.e., exceeding the overlap threshold. Figure 6 This illustrates an overlap where the hand overlaps a sufficiently large area of ​​object 12.

[0068] If overlap exists, the relevant temporary objects previously in the valid class (reference numeral 180) are moved from the valid class to the invalid class (reference numeral 210). Figure 6 In the case shown, object 12 is moved to an invalid class.

[0069] If there is no overlap, for example, in object 12 according to Figure 4 and Figure 5 In the case of object 12, object 12 is retained in the valid class.

[0070] Since in 3D cameras, due to defective pixels or evaluation errors, regions of the image often lack depth values, and meaningful evaluation is impossible when such a region 52 without depth values ​​exists, preferably, the search for overlap also includes searching for overlap between temporary objects in the valid class and the region 52 lacking depth values. If a sufficiently large overlap is determined, the relevant temporary object is moved from the valid class to the invalid class. Such a situation occurs in... Figure 7 It is displayed in the middle.

[0071] Following these steps, in step 220, the temporary object remaining in the valid class is defined as the object to be measured.

[0072] This identifies the start time point and provides a start signal, then in step 230, the volume detection of object 12 is initiated (start time point).

[0073] Volume data is calculated in volume determination unit 20 based on the 3D image of camera 16 and forms part of basic data (which is collected in memory 28), and then output through output unit 22 and output terminal 24.

[0074] If no object 12 remains in the valid class at the time of startup determination, meaning that basic data detection cannot be performed, then outputting an error signal is reasonable. Consequently, no startup signal will be output either.

Claims

1. A method for detecting basic data of an object (12), comprising the following steps: Place the object on the balance (14). The following steps are used to determine the start time point for recording basic data: The camera (16) records an image of the object (12) with depth values. Continuous regions in the image are measured, and these continuous regions then form the first temporary object. The first temporary objects are classified into a valid class and an invalid class. Specifically, all first temporary objects located inside the image and not in contact with the image edge (33) are classified as the valid class, while first temporary objects in contact with the edge are classified as the invalid class. By evaluating the depth value and forming a new sub-region with a depth value exceeding a depth value threshold, the first temporary object in the invalid class is analyzed. The sub-regions are examined to find contiguous regions that form a second temporary object. If the second temporary object does not touch the image edge (33), then the second temporary object is moved to the valid class. Find the overlap between temporary objects in the valid class and temporary objects remaining in the invalid class, wherein the overlap must exceed an overlap threshold. If there is overlap, the relevant temporary object is moved from the valid class to the invalid class. Define the temporary object in the valid class as the object to be measured. If there are no objects in the valid class, an error signal is output; otherwise... After the start time point determined in this manner, volume detection of the object (12) is initiated. Record the aforementioned basic data. Store the basic data that is recorded. Output the basic data.

2. The method according to claim 1, characterized in that, To evaluate the depth values ​​and form new sub-regions with depth values ​​exceeding a depth value threshold, depth value histograms are recorded and evaluated.

3. The method according to claim 1 or 2, characterized in that, In the case of regions lacking depth values, the search for overlap also includes searching for overlaps between temporary objects in the valid class and regions lacking depth values, and if there is an overlap, moving the relevant temporary objects from the valid class to the invalid class.

4. An apparatus for detecting basic data of an object (12), for performing the method according to any one of claims 1 to 3, said apparatus comprising: Balance (14) Camera (16), which is used to record images of the object (12), The evaluation unit (18) has a volume determination unit (20) for determining the volume of the object (12). The initiator unit (26) automatically determines the start time point for recording basic data related to the object's volume when the object (12) is placed on the balance (14). Memory (28), which is used to store the recorded basic data, Output unit (22), which is used to output the basic data, Its features are, The initiator unit (26) is designed to determine the initiation time point based on the image recorded by the camera (16), and for this purpose, the initiator unit (26) has an image evaluation unit (30), and the image evaluation unit (30) is designed to perform object classification, by which foreign objects (50) in addition to the object (12) are also classified as allowed foreign objects or disallowed foreign objects.

5. The device according to claim 4, characterized in that, The initiator unit is designed to output the initiation time point while the allowed external object is still within the image area.

6. The device according to claim 4, characterized in that, The camera is designed to generate images with depth values.

7. The device according to claim 5, characterized in that, The camera is designed to generate images with depth values.

8. The device according to claim 6, characterized in that, The image evaluation unit is designed to perform the object classification while taking into account the depth value.

9. The device according to claim 7, characterized in that, The image evaluation unit is designed to perform the object classification while taking into account the depth value.

10. The device according to any one of claims 4 to 9, characterized in that, The image evaluation unit is designed to output a message if the start time cannot be determined during the waiting time.

Images