Self-checking method for cigarette packet appearance image detection system

Abstract

The invention discloses a self-detection method for a cigarette packet appearance image detection system, and relates to the technical field of image detection, and the method specifically comprises the following steps: 1, carrying out the image collection of a cigarette packet product on a production line through an image collection module, and submitting a collected image to an image processing system; step 2, the image processing system preprocesses a built-in standard data packet to construct a standard preset model, receives the collected image for data analysis, and feeds back an identification result to the control system; and step 3, the control system executes corresponding operation according to the received feedback result. According to the invention, data analysis is carried out on a standard data packet mainly through the image processing system, whether an unrecognizable area exists or not is judged, then whether problems exist in detection function and detection area setting or not is judged, and the reliability of an image detection system is improved; and stable production is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of image detection, and in particular to a self-detection method for a cigarette package appearance image detection system. Background Art

[0002] Cigarette package appearance image detection is widely used to detect packaging material defects, printing errors and other problems to ensure the packaging quality of the product. The principle of cigarette package appearance image detection is to convert the captured target into an image signal through the image acquisition module, and then transmit it to a dedicated image processing system. In the image processing system, the image signal is converted into a digital signal based on pixel distribution and information such as brightness and color. Finally, these digital signals are analyzed by the control module, and various operations are performed to extract the characteristics of the target, and the equipment on site is controlled according to the results of the judgment.

[0003] At present, when cigarette package appearance image detection equipment is used on-site, it is prone to downtime, missing parameters, inaccurate recognition due to external environmental factors, and the execution module may randomly reject cigarette packages, and parameters may be changed without records. If they are not discovered in time by humans, cigarette packages that do not meet the process requirements will flow to the next process, affecting product quality. Therefore, it is necessary to perform image detection on the appearance images of cigarette packages. However, in the existing image detection, it is difficult to monitor the defects of some images with too low resolution, making them impossible to identify, which affects the accuracy of monitoring. For this reason, the present invention provides a self-inspection method for a cigarette package appearance image detection system. Summary of the invention

[0004] In view of the above shortcomings, the present invention provides a self-inspection method for a cigarette package appearance image detection system, which can solve the problems raised in the above background technology.

[0005] In order to achieve the above object, the present invention adopts the following technical scheme:

[0006] A self-checking method for a cigarette package appearance image detection system comprises the following steps:

[0007] Step 1: Capture images of cigarette packages on the production line through the image acquisition module and submit the captured images to the image processing system;

[0008] Step 2: The image processing system preprocesses the built-in standard data packet to build a standard preset model, receives the collected image for data analysis, and feeds the recognition result back to the control system;

[0009] Step 3: The control system performs corresponding operations according to the feedback received: when the feedback result is normal recognition, the control system uploads the image mark to the remote server; if the feedback result is abnormal recognition, the control system records and saves the acquired image, sends out a warning signal, and prompts the technician to conduct manual inspection in time; the control system performs calibration operations on the acquired image after inspection.

[0010] As a preferred implementation, the image labels include correct images, erroneous images and abnormal images.

[0011] As a preferred implementation, the built-in standard data packet in step 2 includes a correct image group and an error image group.

[0012] As a preferred implementation, the pretreatment method of step 2 is as follows:

[0013] D1: The image of the standard data packet is segmented and stored in the standard preset model, and is segmented into D1, D2, D3, ...Di, where i=n, ​​i represents different areas, and n represents the number of different areas.

[0014] As a preferred implementation, the specific manner in which the image processing system performs data analysis on the collected image is as follows:

[0015] S1: Segment the acquired image, that is, divide the image into A1, A2, A3, ... Ai, where i = n, i represents different areas, and n represents the number of different areas;

[0016] S2: Obtain the standard preset model, identify the segmented image with the standard preset model, and determine whether there is an unrecognizable area. If all are correctly identified, the recognition result is normal, and the recognition result is fed back to the control system. The collected image is marked as a correct image or an incorrect image based on the comparison result; if not all are recognized, recognition abnormality analysis is performed.

[0017] As a preferred implementation, the specific steps of identifying abnormal analysis in step S2 include the following:

[0018] When the segmented images of the acquired image are not all correctly identified, the number of regions with abnormal identification is calculated as p, and it is determined whether p exceeds half of the number of segmented regions n, that is:

[0019] when The recognition result is abnormal recognition, the recognition result is fed back to the control system, and the collected image is marked as an abnormal image;

[0020] when Then perform resolution analysis.

[0021] As a preferred embodiment, when the number of abnormal regions is identified The specific steps for further analysis include the following:

[0022] T1: Calculate the resolution values ​​of the segmented images A1, A2, A3, ...Ai as FB1, FB2, FB3, ...FB respectively i , and calculate the average resolution FB 平均 ;

[0023]

[0024] T2: Set the resolution value FB of the segmented image area corresponding to the recognition abnormality i Average value FB with resolution 平均 Make comparisons;

[0025] If FB i ≥FB 平均 , indicating that the segmented image region does not have a problem of being unable to be recognized due to low resolution, that is, there is a problem with the segmented image region itself, and the recognition result is regarded as abnormal recognition, and the recognition result is fed back to the control system, and the collected image is marked as an abnormal image;

[0026] If FB i <FB 平均 , indicating that the segmented image may be unrecognizable due to low resolution, so contrast enhancement analysis is performed.

[0027] As a preferred implementation, the specific steps of contrast enhancement analysis include the following:

[0028] The contrast of the area where the segmented image with too low resolution is located is enhanced, and then the corresponding area is re-compared with the standard preset model. If it can be completely recognized, if all are correctly recognized, the recognition result is normal recognition, and the recognition result is fed back to the control system; if it cannot be recognized, the recognition result is abnormal recognition, and the recognition result is fed back to the control system, and the collected image is marked as an abnormal image.

[0029] As a preferred implementation, the calibration operation of the acquired image after inspection described in step three is specifically as follows: the technician makes a judgment on the acquired image marked as an abnormal image, and if the acquired image of the abnormal image is determined to be correct, it is placed in the correct image group in the standard data package; if the acquired image of the abnormal image is determined to be wrong, it is placed in the wrong image group in the standard data package.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention performs data analysis on the standard data packet through the image processing system, specifically performs image segmentation processing, determines whether there are unrecognizable areas, and then determines whether there are problems with the image, and further analyzes the unrecognizable image through resolution, thereby avoiding the influence of the judgment result due to the image with too low resolution, and improving the accuracy of image detection; the image processing system can be calibrated for parameters to ensure stable production. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for describing the embodiments are briefly introduced below.

[0032] Figure 1 The present invention is a flow chart of a self-checking method for a cigarette package appearance image detection system. DETAILED DESCRIPTION

[0033] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they cannot be understood as limitations on the present invention.

[0035] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" 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 it can be indirectly connected through an intermediate medium, or it can be the internal connection of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0036] The present invention provides a self-test method for a cigarette package appearance image detection system, such as Figure 1 As shown, the specific steps include:

[0037] Step 1: Capture images of cigarette packages on the production line through the image acquisition module and submit the captured images to the image processing system;

[0038] Step 2: The image processing system preprocesses the built-in standard data packet to build a standard preset model, receives the collected image for data analysis, and feeds the recognition result back to the control system;

[0039] Step 3: The control system performs corresponding operations according to the feedback received: when the feedback result is normal recognition, the control system uploads the image mark to the remote server; if the feedback result is abnormal recognition, the control system records and saves the acquired image, sends out a warning signal, and prompts the technician to conduct manual inspection in time; the control system performs calibration operations on the acquired image after inspection.

[0040] The image labels include correct images, erroneous images and abnormal images.

[0041] The built-in standard data packet in step 2 includes a correct image group and an error image group.

[0042] The method for pre-processing in step 2 is as follows:

[0043] D1: The image of the standard data packet is segmented and stored in the standard preset model, and is segmented into D1, D2, D3, ...Di, where i=n, ​​i represents different areas, and n represents the number of different areas.

[0044] The specific method of performing data analysis on the collected image by the image processing system is as follows:

[0045] S1: Segment the acquired image, that is, divide the image into A1, A2, A3, ... Ai, where i = n, i represents different areas, and n represents the number of different areas;

[0046] S2: Obtain the standard preset model, identify the segmented image with the standard preset model, and determine whether there is an unrecognizable area. If all are correctly identified, the recognition result is normal, and the recognition result is fed back to the control system. The collected image is marked as a correct image or an incorrect image based on the comparison result; if not all are recognized, recognition abnormality analysis is performed.

[0047] The specific steps of identifying abnormal analysis in step S2 include the following:

[0048] When the segmented images of the acquired image are not all correctly identified, the number of regions with abnormal identification is calculated as p, and it is determined whether p exceeds half of the number of segmented regions n, that is:

[0049] when The recognition result is abnormal recognition, the recognition result is fed back to the control system, and the collected image is marked as an abnormal image;

[0050] when Then perform resolution analysis.

[0051] The number of regions of the segmented image of the acquired image that are abnormal The specific steps for identifying abnormal analysis include the following:

[0052] T1: Calculate the resolution values ​​of the segmented images A1, A2, A3, ...Ai as FB1, FB2, FB3, ...FB respectively i , and calculate the average resolution FB 平均 ;

[0053]

[0054] T2: Set the resolution value FB of the segmented image area corresponding to the recognition abnormality i Average value FB with resolution 平均 Make comparisons;

[0055] If FB i ≥FB 平均 , indicating that the segmented image region does not have a problem of being unable to be recognized due to low resolution, that is, there is a problem with the segmented image region itself, and the recognition result is regarded as abnormal recognition, and the recognition result is fed back to the control system, and the collected image is marked as an abnormal image;

[0056] If FB i <FB 平均 , indicating that the segmented image may be unrecognizable due to low resolution, so contrast enhancement analysis is performed;

[0057] The specific steps of the contrast enhancement analysis include the following:

[0058] The contrast of the area where the segmented image with too low resolution is located is enhanced, and then the corresponding area is re-identified with the standard preset model. If the recognition can be completed, if all are correctly identified, the recognition result is normal recognition, and the recognition result is fed back to the control system; if the recognition cannot be performed, the recognition result is abnormal recognition, and the recognition result is fed back to the control system, and the collected image is marked as an abnormal image.

[0059] The calibration operation of the acquired image after inspection described in step three is specifically as follows: the technician makes a judgment on the acquired image marked as an abnormal image, and if the acquired image of the abnormal image is judged to be correct, it is placed in the correct image group in the standard data package; if the acquired image of the abnormal image is judged to be wrong, it is placed in the wrong image group in the standard data package.

[0060] In step three, the control system uploads the image mark to the remote server. When the image mark is a correct image, the rejection station passes the current cigarette package product to the next process; when the image mark is an incorrect image, the rejection station rejects the current cigarette package product so that it cannot enter the next process.

[0061] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A self-test method for a cigarette package appearance image detection system, comprising the following steps: Step 1: Capture images of cigarette packages on the production line through the image acquisition module and submit the captured images to the image processing system; Step 2: The image processing system preprocesses the built-in standard data packet to build a standard preset model, receives the collected image for data analysis, and feeds the recognition result back to the control system; Step 3: The control system performs corresponding operations according to the feedback received: when the feedback result is normal recognition, the control system uploads the image mark to the remote server; if the feedback result is abnormal recognition, the control system records and saves the acquired image, sends out a warning signal, and prompts the technician to conduct manual inspection in time; the control system performs calibration operations on the acquired image after inspection.

2. According to a self-inspection method for a cigarette package appearance image detection system according to claim 1, the image marks include correct images, error images and abnormal images.

3. According to the self-test method for a cigarette package appearance image detection system as described in claim 1, the built-in standard data packet in step 2 includes a correct image group and an error image group.

4. According to the self-inspection method for a cigarette package appearance image detection system according to claim 1, the pre-processing method of step 2 is as follows: D1: The image of the standard data packet is segmented and stored in the standard preset model, and is segmented into D1, D2, D3, ...Di, where i=n, ​​i represents different areas, and n represents the number of different areas.

5. According to the self-inspection method for a cigarette package appearance image detection system according to claim 1, the specific manner in which the image processing system performs data analysis on the collected image is as follows: S1: Segment the acquired image, that is, divide the image into A1, A2, A3, ... Ai, where i = n, i represents different areas, and n represents the number of different areas; S2: Obtain the standard preset model, identify the segmented image with the standard preset model, and determine whether there is an unrecognizable area. If all are correctly identified, the recognition result is normal, and the recognition result is fed back to the control system. The collected image is marked as a correct image or an incorrect image based on the comparison result; if not all are recognized, recognition abnormality analysis is performed.

6. According to the self-checking method for a cigarette package appearance image detection system in claim 5, the specific steps of identifying abnormal analysis in step S2 include the following: When the segmented images of the acquired image are not all correctly identified, the number of regions with abnormal identification is calculated as p, and it is determined whether p exceeds half of the number of segmented regions n, that is: when The recognition result is abnormal recognition, the recognition result is fed back to the control system, and the collected image is marked as an abnormal image; when Then perform resolution analysis.

7. The self-checking method for a cigarette package appearance image detection system according to claim 6, when the number of abnormal areas is identified The specific steps for further analysis include the following: T1: Calculate the resolution values ​​of the segmented images A1, A2, A3, ...Ai as FB1, FB2, FB3, ...FB respectively i , and calculate the average resolution FB 平均 ; T2: Set the resolution value FB of the segmented image area corresponding to the recognition abnormality i Average value FB with resolution 平均 Make comparisons; If FB i ≥FB 平均 , indicating that the segmented image region does not have a problem of being unable to be recognized due to low resolution, that is, there is a problem with the segmented image region itself, and the recognition result is regarded as abnormal recognition, and the recognition result is fed back to the control system, and the collected image is marked as an abnormal image; If FB i <FB 平均 , indicating that the segmented image may be unrecognizable due to low resolution, so contrast enhancement analysis is performed.

8. According to the self-test method for a cigarette package appearance image detection system of claim 7, the specific steps of contrast enhancement analysis include the following: The contrast of the area where the segmented image with too low resolution is located is enhanced, and then the corresponding area is re-compared with the standard preset model. If it can be completely recognized, if all are correctly recognized, the recognition result is normal recognition, and the recognition result is fed back to the control system; if it cannot be recognized, the recognition result is abnormal recognition, and the recognition result is fed back to the control system, and the collected image is marked as an abnormal image.

Images