Deep learning-based seal detection method, device and system and storage medium

Abstract

The invention discloses a seal detection method, device and system based on deep learning and a storage medium, and relates to the cigarette packet detection technology, and the method comprises the steps: obtaining cigarette packet images shot by a plurality of cameras; the cigarette packet image shot by each camera is input into a pre-configured deep learning model, the input of the deep learning model is the cigarette packet image, and the output of the deep learning model is a seal detection result of the cigarette packet; and removing abnormal cigarette packets according to the seal detection result. According to the technical scheme, the cigarette packet images can be obtained by shooting the cigarette packet from different directions through the multiple cameras, the cigarette packet images are analyzed through the deep learning models matched with all the directions, the seal detection result is obtained, automatic seal detection is achieved, and the seal detection efficiency is improved. The abnormal cigarette packets are removed according to the seal detection result, the abnormal cigarette packets can be automatically removed, the situation that the abnormal cigarette packets block a channel is avoided, and the cigarette packet production reliability is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of cigarette package detection, and in particular to a seal detection method, device, system and storage medium based on deep learning. Background Art

[0002] In the process of cigarette pack production, cigarette pack seals may be damaged. When cigarette packs with damaged seals enter the cigarette pack channel, the channel will be blocked or flow into the next process.

[0003] At present, the cigarette pack seal status on the production line is checked manually during the cigarette pack production process. How to automatically perform cigarette pack seal inspection and improve the efficiency of cigarette pack seal inspection has become an urgent problem to be solved. Summary of the invention

[0004] The present invention provides a seal detection method, device, system and storage medium based on deep learning to solve the problem of low efficiency of current cigarette package seal detection.

[0005] According to one aspect of the present invention, a seal detection method based on deep learning is provided, comprising:

[0006] Acquire cigarette package images captured by multiple cameras;

[0007] Inputting the cigarette package images captured by each camera into a pre-configured deep learning model, wherein the input of the deep learning model is the cigarette package image and the output is the seal detection result of the cigarette package;

[0008] Abnormal cigarette packs are removed according to the seal detection result.

[0009] According to another aspect of the present invention, a seal detection device based on deep learning is provided, comprising:

[0010] An image acquisition module, used to acquire cigarette package images captured by multiple cameras;

[0011] A model detection module, used to input the cigarette package images taken by each camera into a pre-configured deep learning model, wherein the input of the deep learning model is the cigarette package image and the output is the seal detection result of the cigarette package;

[0012] The rejection module is used to reject abnormal cigarette packs according to the seal detection result.

[0013] According to another aspect of the present invention, a seal detection system based on deep learning is provided, characterized in that it includes:

[0014] Control cabinet, image acquisition components and packaging machine electronic control system;

[0015] The control cabinet includes a power supply component and a control panel component; the power supply component is connected to the control panel component;

[0016] The control board assembly includes at least one processor; and a memory in communication with the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the seal detection method based on deep learning according to any one of claims 1 to 6;

[0017] The image acquisition component includes a plurality of cameras and a plurality of light sources; the control board component is connected to the plurality of cameras and the plurality of light sources;

[0018] The control panel assembly is connected to the electronic control system of the packaging machine.

[0019] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the seal detection method based on deep learning described in any embodiment of the present invention when executed.

[0020] The technical solution of the embodiment of the present invention obtains cigarette package images captured by multiple cameras; the cigarette package images captured by each camera are respectively input into a pre-configured deep learning model, the input of the deep learning model is the cigarette package image, and the output is the seal detection result of the cigarette package; abnormal cigarette packages are eliminated according to the seal detection result. Compared with the current manual seal detection, the technical solution provided by the embodiment of the present invention can obtain cigarette package images by shooting cigarette packages from different directions with multiple cameras, and analyze the cigarette package images using deep learning models adapted to each direction to obtain seal detection results, thereby realizing automated seal detection and improving seal detection efficiency. Abnormal cigarette packages are eliminated according to the seal detection results, which can realize automated elimination of abnormal cigarette packages, avoid situations such as abnormal cigarette packages blocking channels, and improve the reliability of cigarette package production.

[0021] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 It is a flowchart of a seal detection method based on deep learning provided by an embodiment of the present invention;

[0024] Figure 2 is a schematic diagram of camera positions provided by an embodiment of the present invention;

[0025] Figure 3 It is a schematic diagram of the whole process of a seal detection method based on deep learning provided by an embodiment of the present invention;

[0026] Figure 4 is a structural schematic diagram of a seal detection device based on deep learning provided by an embodiment of the present invention;

[0027] Figure 5 It is a structural schematic diagram of a seal detection system based on deep learning provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] In the process of cigarette pack production, cigarette pack seals may be damaged. When cigarette packs with damaged seals enter the cigarette pack channel, the channel will be blocked or flow into the next process.

[0031] At present, the cigarette pack seal status on the production line is checked manually during the cigarette pack production process. How to automatically perform cigarette pack seal inspection and improve the efficiency of cigarette pack seal inspection has become an urgent problem to be solved.

[0032] Figure 1 This is a flow chart of a seal detection method based on deep learning provided by an embodiment of the present invention. This embodiment can be applied to the situation where cigarette pack seals are detected during cigarette pack production. The seal detection device based on deep learning can be implemented in the form of hardware and / or software. The seal detection device based on deep learning can be configured in electronic devices such as control cabinets and computers. Figure 3 As shown, the method is applied to the middleware of the distributed database, including:

[0033] Step S101, obtaining cigarette package images captured by multiple cameras.

[0034] Optionally, obtaining cigarette package images captured by multiple cameras may be implemented in the following manner:

[0035] Multiple cameras are respectively located on both sides of one end and the other end of the cigarette pack conveyor belt, and are used to capture the left view, right view and top view of the cigarette pack respectively; when it is detected that the cigarette pack is placed on the conveyor belt, the multiple cameras are controlled to capture the cigarette pack respectively to obtain the cigarette pack image.

[0036] Figure 2 A schematic diagram of camera positions provided in an embodiment of the present invention, such as Figure 2 As shown, three cameras can be configured, one camera is located at one end of the conveyor belt, and the other two cameras are located on both sides of the other end of the conveyor belt.

[0037] The three cameras take pictures of the cigarette pack from different angles to obtain images of the cigarette pack. The three cameras can obtain images of the cigarette pack from three directions, and then detect the seals at different positions on the cigarette pack from three directions.

[0038] Optionally, before acquiring the cigarette package images captured by multiple cameras, the method further includes:

[0039] Sample data is acquired according to the camera position; labels are added according to the sample data to obtain training data, wherein the training data includes negative samples with abnormal labels and positive samples with normal labels; and an initial deep learning model is trained according to the training data.

[0040] Optionally, adding labels according to the sample data may be implemented as follows:

[0041] An abnormal label is added according to the abnormal situation of the sample data, and the abnormal situation includes: missing seal, damaged seal, offset seal or incomplete seal pattern.

[0042] The above implementation can add abnormal labels to samples according to different abnormal types of seals. Based on the sample data with different types of abnormal labels, training can be performed to determine the specific abnormal type of the cigarette package based on whether the cigarette package is abnormal, thereby improving the accuracy of cigarette package detection.

[0043] The above implementation method can train the initial deep learning model according to sample data at different positions, so that the trained deep learning model can accurately identify abnormal cigarette package images based on the cigarette package images, thereby improving the recognition accuracy.

[0044] Step S102: input the cigarette package images taken by each camera into a pre-configured deep learning model respectively, wherein the input of the deep learning model is the cigarette package image, and the output is the seal detection result of the cigarette package.

[0045] Optionally, the deep learning model can be a convolutional neural network, and the model uses the target detection yolov8 algorithm

[0046] Figure 3 FIG. 1 is a flow chart of a seal detection method based on deep learning provided by an embodiment of the present invention. Figure 3 As shown in the figure, the detection process of the deep learning model includes image acquisition, sample collection, sample classification, model creation, model training and model deployment. After the model is deployed, the following steps are involved in using the model: converting the model format, introducing dynamic library connections, reading target files, loading model files, converting images to tensors, model input, executing the inference network, obtaining results and model optimization and update.

[0047] Step S101 uses three cameras to obtain cigarette pack images respectively, and each camera has a corresponding deep learning model for detecting anomalies in the cigarette pack images captured by the camera. After training, the deep learning model can determine whether there is a seal anomaly in the cigarette pack image based on the cigarette pack image.

[0048] Three deep learning models can be set up to recognize the left view, right view, and top view respectively.

[0049] Step S103: rejecting abnormal cigarette packs according to the seal detection result.

[0050] Optionally, rejecting abnormal cigarette packages according to the seal detection result can be implemented in the following manner:

[0051] If the seal detection result is abnormal, a rejection signal is output to the electronic control system of the packaging machine so that the electronic control system of the packaging machine rejects the cigarette package through the packaging unit; if the seal detection result is normal, the rejection signal output to the electronic control system of the packaging machine is cancelled.

[0052] When the seal detection result is abnormal, a rejection signal is output to the electronic control system of the packaging machine. The electronic control system of the packaging machine is used to control the transmission and rejection of cigarette packs. The cigarette pack rejection action can be performed through the packaging unit.

[0053] Furthermore, after removing abnormal cigarette packages according to the seal detection result, the method further includes:

[0054] Perform an accuracy check based on the seal detection result; if the check fails, obtain a supplementary training sample; and train the deep learning model based on the supplementary training sample.

[0055] After the deep learning model is deployed, the detection result of the deep learning model is verified by inspection. If the detection result of the deep learning model is found to be inaccurate through verification, the deep learning model is supplemented with training. Specifically, supplementary training samples are obtained, and the deep learning model is trained according to the supplementary training samples.

[0056] The seal detection method based on deep learning provided in the embodiment of the present invention obtains cigarette package images captured by multiple cameras; the cigarette package images captured by each camera are respectively input into a pre-configured deep learning model, the input of the deep learning model is the cigarette package image, and the output is the seal detection result of the cigarette package; abnormal cigarette packages are eliminated according to the seal detection result. Compared with the current manual seal detection, the seal detection method based on deep learning provided in the embodiment of the present invention can obtain cigarette package images by shooting cigarette packages from different directions with multiple cameras, and analyze the cigarette package images using deep learning models adapted to each direction to obtain seal detection results, thereby realizing automated seal detection and improving seal detection efficiency. Abnormal cigarette packages are eliminated according to the seal detection results, which can realize automated elimination of abnormal cigarette packages, avoid situations such as abnormal cigarette packages blocking channels, and improve the reliability of cigarette package production.

[0057] Figure 4 is a structural schematic diagram of a seal detection device based on deep learning provided by an embodiment of the present invention. This embodiment can be applied to the situation where cigarette pack seals are detected during cigarette pack production, such as Figure 4 As shown, the device comprises:

[0058] An image acquisition module 21 is used to acquire cigarette package images captured by multiple cameras;

[0059] A model detection module 22, used to input the cigarette package images taken by each camera into a pre-configured deep learning model, wherein the input of the deep learning model is the cigarette package image, and the output is the seal detection result of the cigarette package;

[0060] The rejection module 23 is used to reject abnormal cigarette packs according to the seal detection result.

[0061] On the basis of the above embodiment, optionally, a training module is further included, which is used to obtain sample data according to the camera position before obtaining the cigarette package images captured by multiple cameras;

[0062] Adding labels according to the sample data to obtain training data, wherein the training data includes negative samples with abnormal labels and positive samples with normal labels;

[0063] The initial deep learning model is trained according to the training data.

[0064] Based on the above embodiment, optionally, the training module is used to:

[0065] An abnormal label is added according to the abnormal situation of the sample data, and the abnormal situation includes: missing seal, damaged seal, offset seal or incomplete seal pattern.

[0066] Based on the above embodiment, optionally, the image acquisition module 21 is used to:

[0067] Multiple cameras are located on both sides of one end and the other end of the cigarette pack conveyor belt, respectively, for photographing the left view, right view and top view of the cigarette pack;

[0068] When it is detected that a cigarette pack is placed on the conveyor belt, the multiple cameras are controlled to respectively photograph the cigarette pack to obtain a cigarette pack image.

[0069] Based on the above embodiment, optionally, the rejection module 23 is used to:

[0070] If the seal detection result is abnormal, a rejection signal is output to the electronic control system of the packaging machine so that the electronic control system of the packaging machine rejects the cigarette package through the packaging unit;

[0071] If the seal detection result is normal, the rejection signal output to the electronic control system of the packaging machine is cancelled.

[0072] Based on the above embodiment, optionally, the training module is further used to perform accuracy verification according to the seal detection result after the abnormal cigarette packs are removed according to the seal detection result;

[0073] If the verification fails, obtain additional training samples;

[0074] The deep learning model is trained according to the supplementary training samples.

[0075] The seal detection device based on deep learning provided in the embodiment of the present invention comprises an image acquisition module 21 for acquiring cigarette package images captured by multiple cameras; a model detection module 22 for inputting the cigarette package images captured by each camera into a pre-configured deep learning model, wherein the input of the deep learning model is the cigarette package image and the output is the seal detection result of the cigarette package; and a rejection module 23 for rejecting abnormal cigarette packages according to the seal detection result. Compared with the current manual seal detection, the seal detection device based on deep learning provided in the embodiment of the present invention can obtain cigarette package images by photographing cigarette packages from different directions with multiple cameras, and analyze the cigarette package images using the deep learning models adapted to each direction, to obtain the seal detection results, thereby realizing automated seal detection and improving the seal detection efficiency. By rejecting abnormal cigarette packages according to the seal detection results, it is possible to realize automated rejection of abnormal cigarette packages, avoid situations such as abnormal cigarette packages blocking the channel, and improve the reliability of cigarette package production.

[0076] The deep learning-based seal detection device provided in the embodiment of the present invention can execute the deep learning-based seal detection method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0077] Figure 5 is a structural schematic diagram of a seal detection system based on deep learning provided by an embodiment of the present invention, the system comprising:

[0078] Control cabinet 31, image acquisition component 32 and packaging machine electronic control system 33;

[0079] The control cabinet 31 includes a power supply component and a control panel component; the power supply component is connected to the control panel component;

[0080] The control board assembly includes at least one processor; and a memory in communication with the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the seal detection method based on deep learning according to any one of claims 1 to 6;

[0081] The image acquisition component 32 includes a plurality of cameras and a plurality of light sources; the control board component is connected to the plurality of cameras and the plurality of light sources;

[0082] The control panel assembly is connected to the packaging machine electrical control system 33 .

[0083] On the basis of the above embodiment, optionally, the control cabinet 31 further includes a visual industrial computer and a human-computer interaction group component, the human-computer interaction group component is connected to the visual industrial computer, and the visual industrial computer is also connected to the control board component; the power supply component is respectively connected to the visual industrial computer and the human-computer interaction group component;

[0084] The human-computer interaction group component includes a display screen; the visual industrial computer is connected to the multiple cameras and multiple light sources.

[0085] Optionally, the visual industrial computer is connected to the display component of the human-computer interaction group component through a display interface. The visual industrial computer is connected to the RS422 interface of the control board component through an RS422 interface, and the processor of the control board component is connected to multiple light sources and cameras through a light source and camera trigger interface. Among them, the number of light sources and cameras is three. The image acquisition component 32 is composed of 3 groups of LED light sources, 3 groups of Ethernet cameras, 3 groups of low-distortion industrial lenses and corresponding mounting structures, and is mainly used to collect the end face image of the cigarette package seal.

[0086] The control cabinet 31 synchronizes the phase with the packaging machine host through the encoder. At a fixed phase, the control cabinet 31 controls the image acquisition component 32 to collect the end face image of the cigarette package seal. The industrial computer calculates and processes the image after receiving it. If the result is abnormal, the control cabinet 31 sends a rejection signal to the packaging machine control system, and the packaging unit rejects the cigarette package at the corresponding position. If the result is normal, the control cabinet 31 does not output a signal.

[0087] The seal detection system based on deep learning includes at least one processor, and a memory connected to the at least one processor, such as a read-only memory (ROM), a random access memory (RAM), etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) or the computer program loaded from the storage unit to the random access memory (RAM). In RAM, various programs and data required for the operation of electronic devices can also be stored. The processor, ROM and RAM are connected to each other via a bus. The input / output (I / O) interface 15 is also connected to the bus.

[0088] The processor may be any general and / or special processing component with processing and computing capabilities. Some examples of processors include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the seal detection method based on deep learning.

[0089] In some embodiments, the seal detection method based on deep learning can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit. In some embodiments, part or all of the computer program can be loaded and / or installed on an electronic device via a ROM and / or a communication unit. When the computer program is loaded into RAM 13 and executed by the processor, one or more steps of the seal detection method based on deep learning described above can be performed. Alternatively, in other embodiments, the processor can be configured to perform the seal detection method based on deep learning in any other appropriate manner (for example, by means of firmware).

[0090] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0091] The computer programs for implementing the deep learning-based seal detection method of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program can be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0092] The embodiment of the invention further provides a computer-readable storage medium, the computer-readable storage medium storing computer instructions, the computer instructions being used to enable a processor to execute a seal detection method based on deep learning, applied to middleware of a distributed database, including:

[0093] Acquire cigarette package images captured by multiple cameras;

[0094] Inputting the cigarette package images captured by each camera into a pre-configured deep learning model, wherein the input of the deep learning model is the cigarette package image and the output is the seal detection result of the cigarette package;

[0095] Abnormal cigarette packs are removed according to the seal detection result.

[0096] Based on the above embodiment, optionally, before acquiring the cigarette package images captured by multiple cameras, the method further includes:

[0097] Get sample data according to the camera position;

[0098] Add labels according to the sample data to obtain training data, wherein the training data includes negative samples with abnormal labels and positive samples with normal labels;

[0099] The initial deep learning model is trained according to the training data.

[0100] Based on the above embodiment, optionally, adding a label according to the sample data includes:

[0101] An abnormal label is added according to the abnormal situation of the sample data, and the abnormal situation includes: missing seal, damaged seal, offset seal or incomplete seal pattern.

[0102] Based on the above embodiment, optionally, obtaining cigarette package images captured by multiple cameras includes:

[0103] Multiple cameras are located on both sides of one end and the other end of the cigarette pack conveyor belt, respectively, for photographing the left view, right view and top view of the cigarette pack;

[0104] When it is detected that a cigarette pack is placed on the conveyor belt, the multiple cameras are controlled to respectively photograph the cigarette pack to obtain a cigarette pack image.

[0105] On the basis of the above embodiment, optionally, rejecting abnormal cigarette packages according to the seal detection result includes:

[0106] If the seal detection result is abnormal, a rejection signal is output to the electronic control system of the packaging machine so that the electronic control system of the packaging machine rejects the cigarette package through the packaging unit;

[0107] If the seal detection result is normal, the rejection signal output to the electronic control system of the packaging machine is cancelled.

[0108] Based on the above embodiment, optionally, after removing abnormal cigarette packs according to the seal detection result, the method further includes:

[0109] Performing accuracy verification based on the seal detection result;

[0110] If the verification fails, obtain additional training samples;

[0111] The deep learning model is trained according to the supplementary training samples.

[0112] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0113] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0114] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0115] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0116] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0117] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A seal detection method based on deep learning, characterized in that: include: Acquire cigarette package images captured by multiple cameras; Inputting the cigarette package images captured by each camera into a pre-configured deep learning model, wherein the input of the deep learning model is the cigarette package image and the output is the seal detection result of the cigarette package; Abnormal cigarette packs are removed according to the seal detection result.

2. The method according to claim 1, characterized in that Before obtaining the cigarette package images captured by multiple cameras, the following steps are also included: Get sample data according to the camera position; Add labels according to the sample data to obtain training data, wherein the training data includes negative samples with abnormal labels and positive samples with normal labels; The initial deep learning model is trained according to the training data.

3. The method according to claim 2, characterized in that Adding labels according to the sample data includes: An abnormal label is added according to the abnormal situation of the sample data, and the abnormal situation includes: missing seal, damaged seal, offset seal or incomplete seal pattern.

4. The method according to claim 1, characterized in that: Get cigarette pack images captured by multiple cameras, including: Multiple cameras are located on both sides of one end and the other end of the cigarette pack conveyor belt, respectively, for photographing the left view, right view and top view of the cigarette pack; When it is detected that a cigarette pack is placed on the conveyor belt, the multiple cameras are controlled to respectively photograph the cigarette pack to obtain a cigarette pack image.

5. The method according to claim 1, characterized in that: Removing abnormal cigarette packages according to the seal detection result includes: If the seal detection result is abnormal, a rejection signal is output to the electronic control system of the packaging machine so that the electronic control system of the packaging machine rejects the cigarette package through the packaging unit; If the seal detection result is normal, the rejection signal output to the electronic control system of the packaging machine is cancelled.

6. The method according to claim 1, characterized in that After the abnormal cigarette packs are removed according to the seal detection result, the method further includes: Performing accuracy verification based on the seal detection result; If the verification fails, obtain additional training samples; The deep learning model is trained according to the supplementary training samples.

7. A seal detection device based on deep learning, characterized in that: include: An image acquisition module, used to acquire cigarette package images captured by multiple cameras; A model detection module, used to input the cigarette package images taken by each camera into a pre-configured deep learning model, wherein the input of the deep learning model is the cigarette package image and the output is the seal detection result of the cigarette package; The rejection module is used to reject abnormal cigarette packs according to the seal detection result.

8. A seal detection system based on deep learning, characterized in that: include: Control cabinet, image acquisition components and packaging machine electronic control system; The control cabinet includes a power supply component and a control panel component; the power supply component is connected to the control panel component; The control board assembly includes at least one processor; and a memory in communication with the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the seal detection method based on deep learning according to any one of claims 1 to 6; The image acquisition component includes a plurality of cameras and a plurality of light sources; the control board component is connected to the plurality of cameras and the plurality of light sources; The control panel assembly is connected to the electronic control system of the packaging machine.

9. The system according to claim 8, characterized in that The control cabinet further comprises a visual industrial computer and a human-computer interaction group component, wherein the human-computer interaction group component is connected to the visual industrial computer, and the visual industrial computer is also connected to the control panel component; the power supply component is respectively connected to the visual industrial computer and the human-computer interaction group component; The human-computer interaction group component includes a display screen; the visual industrial computer is connected to the multiple cameras and multiple light sources.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the seal detection method based on deep learning described in any one of claims 1 to 6 when executed.

Images