Stacking positioning identification method, stacking automatic feeding device and cigarette packaging supply line

Abstract

The invention belongs to the technical field of tobacco processing, and discloses a stacking positioning recognition method, a stacking automatic feeding device and a cigarette packaging supply line. The stacking positioning identification method comprises the following steps: S1, pre-generating a template image and inputting the template image into a database; s2, shooting a material field source image; s3, pre-matching the source image with a template image in a database, and screening out a first image group; and S4, the source image is accurately matched with the first image group, and accurate stacking information is obtained. By pre-generating the template image, the time for generating the template image in the image matching process is shortened; the range of the template image needing to be matched is narrowed through pre-matching, and then the source image is accurately matched, so that the image matching efficiency is improved, and the tobacco production and processing efficiency is improved.

Description

Technical Field

[0001] The invention relates to the technical field of tobacco processing, and in particular to a palletizing positioning and identification method, a palletizing automatic feeding device and a cigarette packaging supply line. Background Art

[0002] Traditional manual cargo handling is gradually being replaced by automation and mechanization, so automatic palletizing has become a common need in the field of industrial production. In the prior art, the specific information of palletizing is identified by image matching. The image matching process is: through a sliding window method, the template image is traversed on the source image, and the pixel matching degree of the overlapping sub-image of the template image and the source image is calculated to obtain the specific information of the palletizing.

[0003] However, since palletizing is done manually, the angles of each pallet will be different. Also, due to the perspective principle of multiple layers of palletizing and the size of actual images of palletizing are different, so a fixed template image will inevitably fail to recognize it. In order to adapt to the multiple angles and scales of palletizing, the template image also needs to be rotated and scaled before matching with the source image, resulting in low efficiency of image matching and affecting production efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a palletizing positioning and recognition method, a palletizing automatic feeding device and a cigarette packaging supply line, so as to solve the problem in the prior art that the template image needs to be rotated and scaled before matching with the source image, resulting in low efficiency of image matching and affecting production efficiency, and to improve the efficiency of palletizing positioning and recognition.

[0005] To achieve this object, the present invention adopts the following technical solutions:

[0006] On the one hand, a palletizing positioning and identification method is provided, comprising the following steps:

[0007] S1, pre-generate template image and input into database;

[0008] S2, taking source images of materials on site;

[0009] S3, pre-matching the source image with the template image in the database to filter out a first image group;

[0010] S4. Accurately match the source image with the first image group to obtain accurate information of the palletizing.

[0011] As an optional technical solution for the palletizing positioning and recognition method, in step S1, pre-generating the template image includes the following steps:

[0012] S1.1, generating a basic first image according to the palletizing of different models;

[0013] S1.2, rotating the generated first image at all angles and saving the image freeze frame at each angle to form a second image;

[0014] S1.3, scaling each of the second images to full scale and saving each scaled image to form a third image;

[0015] S1.4. The first image, the second image and the third image constitute the template image.

[0016] As an optional technical solution for the palletizing positioning and recognition method, in step S3, the source image is reduced and pre-matched with the template image.

[0017] As an optional technical solution for the palletizing positioning and recognition method, in the step S4, the source image is restored to its original size and then accurately matched with the first image group.

[0018] As an optional technical solution of the palletizing positioning and identification method, step S0 is arranged between step S2 and step S3;

[0019] S0. Setting a recognition height range, and screening out the position of the pallet in the source image within the height range.

[0020] As an optional technical solution of the palletizing positioning and identification method, step S00 is set between step S2 and step S3;

[0021] S00, setting logo recognition, and screening out a second image group corresponding to the model of the pallet.

[0022] On the other hand, there is provided an automatic palletizing loading device, which applies any of the above-mentioned palletizing positioning and identification methods, and comprises a robotic arm, wherein a clamp is provided at the free end of the robotic arm, and the clamp can clamp or release the pallet.

[0023] As an optional technical solution for the palletizing automatic loading device, the palletizing automatic loading device further includes a camera, which is detachably mounted on the fixture to obtain the source image of the material site.

[0024] As an optional technical solution for the automatic palletizing loading device, the camera is configured as a 3D camera.

[0025] On the other hand, a cigarette packaging supply line is provided, comprising the above-mentioned automatic palletizing loading device.

[0026] Beneficial effects of the present invention:

[0027] The present invention provides a palletizing positioning and identification method, a palletizing automatic feeding device and a cigarette packaging supply line, the palletizing positioning and identification method comprising the following steps: S1, pre-generating a template image and inputting it into a database; S2, shooting a source image of a material site; S3, pre-matching the source image with the template image in the database, screening out a first image group; S4, accurately matching the source image with the first image group, and obtaining accurate information of the palletizing. By pre-generating a template image, the time for generating a template image in the image matching process is reduced; by pre-matching, the range of the template image to be matched is narrowed, and then the source image is accurately matched, thereby improving the efficiency of image matching and the efficiency of tobacco production and processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a flow chart of a palletizing positioning and identification method provided by an embodiment of the present invention;

[0029] Figure 2 It is a flow chart of a pre-generated template image of a palletizing positioning and recognition method provided by an embodiment of the present invention;

[0030] Figure 3 This is a schematic diagram of a partial structure of the automatic palletizing loading device provided by an embodiment of the present invention. Figure 1 ;

[0031] Figure 4 This is a schematic diagram of a partial structure of the automatic palletizing loading device provided by an embodiment of the present invention. Figure 2 .

[0032] In the figure:

[0033] 100. Palletizing; 10. Robotic arm; 20. Clamp. DETAILED DESCRIPTION

[0034] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0035] In the description of the present invention, unless otherwise clearly specified and limited, the terms "connected", "connected", and "fixed" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0036] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may include that the first and second features are in direct contact, or may include that the first and second features are not in direct contact but are in contact through another feature between them. Moreover, a first feature being "above", "above" and "above" a second feature includes that the first feature is directly above and obliquely above the second feature, or simply indicates that the first feature is higher in level than the second feature. A first feature being "below", "below" and "below" a second feature includes that the first feature is directly below and obliquely below the second feature, or simply indicates that the first feature is lower in level than the second feature.

[0037] In the description of this embodiment, the terms "upper", "lower", "right", etc., directions or positional relationships are based on the directions or positional relationships shown in the drawings, and are only for the convenience of description and simplification of operation, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are only used to distinguish in the description and have no special meaning.

[0038] Traditional manual cargo handling is gradually being replaced by automation and mechanization, so automatic palletizing has become a common need in the field of industrial production. The inventor found that there are three problems in the development of palletizing recognition: 1. Camera installation method; 2. Recognition algorithm selection; 3. Recognition algorithm optimization.

[0039] Regarding question 1, there are two ways to install cameras in visual systems in the prior art:

[0040] Eye-to-hand: The camera is installed outside the robot arm, fixed relative to the robot base (world coordinate system), and does not move with the movement of the robot arm.

[0041] Eye-in-hand: The camera is mounted on a robotic arm and moves with the arm.

[0042] The two methods have different advantages and disadvantages:

[0043]

[0044]

[0045] Since the eye-on-the-hand installation requires the installation of a camera bracket, it has higher requirements for site space and will affect equipment maintenance, etc., and cannot be applied to various different site conditions; and the eye-on-the-hand installation distance is long, the depth of field varies greatly, and the requirements for the camera are very high. Moreover, due to the distortion of the camera lens and other reasons, it is easy to cause recognition errors. Therefore, this embodiment chooses to install the camera in the eye-on-the-hand style.

[0046] In response to question 1, this embodiment provides a cigarette packaging supply line, including a palletizing automatic loading device. Figure 3 and Figure 4 , including a mechanical arm 10, a clamp 20 is provided at the free end of the mechanical arm 10, and the clamp 20 can clamp or release the pallet 100. Further, the palletizing automatic feeding device also includes a camera, which is provided on the clamp 20 to obtain a source image of the material site. In this embodiment, the camera is set as a 3D camera to obtain three-dimensional point cloud data of each position in the source image, so as to increase the accuracy of image recognition.

[0047] For question 2, there are two material recognition algorithms in the prior art:

[0048] Machine learning method: The depalletizing and stacking algorithm based on machine learning must collect a large amount of image data in the early stage, and then do manual labeling, and then go through a long period of training to achieve the ideal recognition effect and accuracy. If there are many types of materials, the learning cycle is very long. If the type of goods being transported is changed, retraining must also be carried out. Therefore, it is only applicable to scenarios with relatively fixed material types.

[0049] Image matching method: The image matching process is matched through a sliding window. On the image to be detected, the image is traversed from left to right and from top to bottom, and the pixel matching degree between the image template and the overlapping sub-image of the source image is calculated. The greater the degree of matching, the greater the possibility that the sub-image is the same as the template. Although the image matching method does not require pre-learning, the image matching method is based on two-dimensional images for cargo identification. Since palletizing is manually stacked, the angles of each pallet will be different. In addition, due to the perspective principle of multi-layer stacking of pallets, the actual image size of the pallet is different, so the fixed template image will inevitably fail to recognize. In order to adapt to the multi-angle and multi-scale of palletizing, the template image also needs to be rotated and scaled before matching with the source image, so the image matching speed is slow, affecting the production rhythm.

[0050] For palletizing recognition, although the type of palletizing to be handled is basically fixed, the types of palletizing in different factories and at different times will vary. If machine learning methods are used, once the material is changed, it will take a long time to re-learn and train, which will affect the work progress. Although the image matching method has a slow matching speed, the overall rhythm of the processing process in the application is not fast (a stack of 250 sheets of paper can meet the production needs of more than 3 minutes), so the image matching method is more suitable for the needs of this application.

[0051] Regarding question 3, although the image matching method is suitable for the needs of this application, there is still room for optimization to further improve the efficiency of image matching.

[0052] Specifically, this embodiment provides a method for palletizing positioning and identification, see Figure 1 , including the following steps:

[0053] S1. Pregenerate template image and input it into the database.

[0054] S2. Take source images of materials on site.

[0055] S3. Pre-match the source image with the template image in the database to select a first image group.

[0056] S4. Accurately match the source image with the first image group to obtain accurate information of the pallet 100.

[0057] By pre-generating template images, the time for generating template images during image matching is reduced; by pre-matching, the range of template images that need to be matched is narrowed, and then the source image is accurately matched, which improves the efficiency of image matching and the efficiency of tobacco production and processing.

[0058] Specifically, see Figure 2 In step S1, pre-generating a template image includes the following steps:

[0059] S1.1. Generate a basic first image according to different types of pallets 100.

[0060] S1.2, rotating the generated first image at all angles and saving the image freeze at each angle to form a second image.

[0061] S1.3, scaling each second image to full scale and saving the image freeze at each scale to form a third image.

[0062] S1.4. The first image, the second image and the third image constitute a template image.

[0063] Furthermore, in step S3, the source image is reduced and pre-matched with the template image. In step S4, the source image is restored to its original size and accurately matched with the first image group. Since image matching requires the use of template images to traverse the source image, the larger the image, the slower the matching speed and the lower the efficiency. Therefore, in the pre-matching step S3, the source image is reduced, and the template image is used to quickly match the reduced source image, and the template image (first image group) that roughly matches the actual image of the pallet 100 in the source image is screened out, which not only improves the matching time, but also reduces the number of template images that need to be matched twice. In the accurate matching step S4, only the template image in the first image group needs to be matched with the source image.

[0064] In other embodiments, in the pre-matching step S3, the template image having the same inclination angle as the actual image of the pallet in the source image may be matched preferentially to reduce the comparison items so as to improve the matching speed; in the precise matching step S4, the template image having the same scaling ratio as the actual image may be compared again in the template image matching the inclination angle of the actual image.

[0065] Furthermore, between step S2 and step S3, step S0 is set; S0, setting the recognition height range, screening out the position of the palletizer 100 within the height range in the source image. Since the palletizers 100 are stacked, a two-dimensional image of the palletizer 100 will appear in the picture taken by the camera. In the process of comparing the template image with the source image, it is also necessary to identify which two-dimensional image of the palletizer 100 is the two-dimensional image of the target palletizer 100 (the top layer of the palletizer), resulting in a reduction in matching speed. The three-dimensional data obtained by the 3D camera corresponds to each point in the two-dimensional image in the source image. By setting the height range, a part of the source image can be excluded, that is, the image matching efficiency is improved by reducing the size of the source image, and the interference of the bottom layer of the palletizer 100 is shielded, thereby improving the accuracy of the image.

[0066] Between step S2 and step S3, step S00 is set; S00, mark recognition is set to filter out the second image group corresponding to the model of the palletizer 100. Since there are different models of palletizers 100, model recognition can be performed before pre-matching step S3 to reduce the number of template images and increase the speed of image matching. Specifically, different models of palletizers 100 can use traditional Chinese characters to increase recognition points and avoid recognition errors. In other embodiments, a circle, concentric circle, or other mark that does not change shape with rotation can be selected to avoid the mark not being recognized after the palletizer 100 is rotated.

[0067] After the model of the pallet 100 is identified, the pre-matching can select an image template to match in the second image group, thereby improving the efficiency of image matching.

[0068] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, various obvious changes, readjustments and substitutions can be made without departing from the protection scope of the present invention. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A palletizing positioning and identification method, characterized in that: The following steps are involved: S1, pre-generate template image and input into database; S2, taking source images of materials on site; S3, pre-matching the source image with the template image in the database to filter out a first image group; S4. Accurately match the source image with the first image group to obtain accurate information of the pallet (100).

2. The method for palletizing positioning and identification according to claim 1, characterized in that: In the step S1, pre-generating the template image comprises the following steps: S1.1, generating a basic first image according to different models of the pallet (100); S1.2, rotating the generated first image at all angles and saving the image freeze frame at each angle to form a second image; S1.3, scaling each of the second images to full scale and saving each scaled image to form a third image; S1.

4. The first image, the second image and the third image constitute the template image.

3. The method for palletizing positioning and identification according to claim 1, characterized in that: In the step S3, the source image is reduced in size and then pre-matched with the template image.

4. The method for palletizing positioning and identification according to claim 3, characterized in that: In the step S4, the source image is restored to its original size and then accurately matched with the first image group.

5. The method for palletizing positioning and identification according to claim 1, characterized in that: Between step S2 and step S3, step S0 is arranged; S0. Setting a recognition height range, and screening out the position of the pallet (100) within the height range in the source image.

6. The method for palletizing positioning and identification according to claim 1, characterized in that: Between step S2 and step S3, step S00 is arranged; S00, setting up logo recognition, and screening out a second image group corresponding to the model of the pallet (100).

7. A palletizing automatic feeding device, using the palletizing positioning and identification method according to any one of claims 1 to 6, characterized in that: It comprises a mechanical arm (10), wherein a clamp (20) is provided at the free end of the mechanical arm (10), and the clamp (20) can clamp or release the pallet (100).

8. The automatic palletizing feeding device according to claim 7, characterized in that: The palletizing automatic loading device also includes a camera, which is arranged on the clamp (20) to obtain the source image of the material site.

9. The automatic palletizing feeding device according to claim 8, characterized in that: The camera is configured as a 3D camera.

10. A cigarette packaging supply line, characterized in that: It comprises the automatic palletizing loading device as described in any one of claims 7 to 9.

Images