Cigarette stain detection system based on deep learning
Through image processing and deep learning smoke roll stain detection systems, the problem of low efficiency and low accuracy of traditional detection methods is solved, and high -precision, automation and real -time smoke roll stain detection are achieved, which improves the detection efficiency and accuracy.
Patent Information
- Application Number
- CN202411452738.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-17
- Publication Date
- 2025-01-24
AI Technical Summary
Traditional cigarette roll quality detection methods rely on manual inspection, are inefficient and easily affected by human factors, making it difficult to ensure detection accuracy and consistency.
Using image processing and deep learning cigarette roll stain detection systems, collecting cigarette roll images through high -resolution cameras, using pre -trained convolutional neural networks (CNN) models for image processing and stain recognition to achieve automated and real -time stain detection and positioning. Essence
It improves the efficiency and accuracy of cigarette appearance quality inspection, achieves high-precision stain detection, reduces manual errors, has automated and real-time detection capabilities, and provides image annotation and alarm functions to support data storage and traceability.
Smart Images
Figure CN119359673A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of cigarette quality detection, and in particular to a cigarette stain detection system based on image processing and deep learning. Background Art
[0002] In the tobacco production process, the appearance quality of cigarettes directly affects the market performance of the product and consumer satisfaction. The surface of cigarettes may have stains or other defects due to various factors in the production, storage, transportation and other links. Traditional quality inspection methods mainly rely on manual inspection, which is not only inefficient, but also easily affected by human factors, and the detection accuracy and consistency are difficult to guarantee. Therefore, there is an urgent need for a technical means to automatically, quickly and accurately detect stains on the surface of cigarettes.
[0003] In recent years, with the rapid development of deep learning and computer vision technology, automated detection systems based on image processing have been widely used in many fields. By introducing deep learning algorithms, the location of stains on the surface of cigarettes can be effectively identified, solving the shortcomings of traditional detection methods and thus improving detection efficiency and accuracy. Summary of the invention
[0004] 1. Purpose of the Invention The purpose of the present invention is to provide a cigarette stain detection system based on deep learning, which can automatically capture cigarette images and process image data through deep learning algorithms to achieve accurate detection and positioning of cigarette stains, thereby improving the efficiency and accuracy of cigarette appearance quality detection.
[0005] 2. Technical solution To achieve the above object, the present invention provides a cigarette stain detection system, comprising the following main components: Image acquisition module: used to acquire images of the cigarette surface, preferably using a high-resolution camera to ensure that the captured images have sufficient details.
[0006] Deep learning processing module: processes the collected images and uses a pre-trained convolutional neural network (CNN) model to identify stains on the surface of cigarettes. This module can automatically extract image features and determine whether there are stains on the surface of cigarettes based on the output of the model.
[0007] Stain Location Module: Based on the output of the deep learning processing module, the identified stain area is annotated and an image output with a stain location box is generated. This module can generate a rectangular or polygonal box to accurately identify the location of the stain.
[0008] Display and alarm module: When the area or quantity of detected stains exceeds the preset threshold, the system will automatically trigger an alarm and display a marked image on the display for the operator's reference.
[0009] Data storage module: This module is used to save the images and stain location information generated during the detection process to facilitate subsequent analysis and quality control.
[0010] 3. Beneficial effects The cigarette stain detection system provided by the present invention has the following beneficial effects: High-precision stain detection: Using deep learning models for image analysis can accurately identify stains on the surface of cigarettes, avoiding errors and inconsistencies in manual detection.
[0011] Automated and real-time detection: The system can realize real-time and automated stain detection during the production process, greatly improving the detection efficiency.
[0012] Image annotation and alarm function: By automatically marking the location of stains, operators can quickly understand where the problem lies; the system can also issue alarms based on preset thresholds to indicate situations that require immediate attention.
[0013] Data storage and tracing: The system has data storage function, which can save the detection images and stain information to facilitate production records and problem tracing. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] Figure 1 : is the overall block diagram of the system of the present invention. This figure shows the relationship and data flow between the main modules. DETAILED DESCRIPTION In order to better understand the present invention, the present invention is now described in detail with reference to the accompanying drawings and specific embodiments: System Structure
[0016] like Figure 1 As shown, the cigarette stain detection system of the present invention is mainly composed of an image acquisition module, a deep learning processing module, a stain positioning module, a display and alarm module and a data storage module.
[0017] Image acquisition module: This module preferably uses a high-resolution industrial camera to capture high-definition images of the cigarette surface in real time on the production line. Through reasonable light source settings, ensure that the details of the cigarette surface can be clearly displayed in the image.
[0018] Deep learning processing module: The system uses a pre-trained convolutional neural network model that is trained on a large number of labeled cigarette stain datasets. In actual applications, the system can extract features from the captured images and classify and locate the stains.
[0019] Stain Localization Module: After the deep learning processing module completes the image analysis, the stain localization module will generate a rectangular or polygonal box in the image based on the analysis results, accurately mark the location of the stain, and output an annotated image.
[0020] Data storage module: All inspection images and their stain location information will be saved in the database for subsequent quality analysis and problem tracing.
[0021] 2. System workflow The system's workflow is as follows: Image acquisition: When the cigarette passes through the conveyor belt, the image acquisition module automatically captures its surface image and transmits the image to the deep learning processing module.
[0022] Image processing and stain detection: The deep learning processing module processes the image, identifies stains through a neural network model, and transmits the location information of the stains to the stain location module.
[0023] Stain annotation: The stain localization module generates a bounding box in the image based on the recognition results and marks the location of the stain.
[0024] Display and alarm: The system displays the labeled image on the display screen. If the detected stains exceed the preset threshold, the system will trigger an alarm.
[0025] Data storage: Test results are automatically saved for subsequent analysis.
[0026] 3. Training Dataset The system uses a large training data set containing images of cigarettes with various stain types and different lighting conditions. By training on these data sets, the system's deep learning model can recognize stains of different types and sizes with high generalization ability.
[0027] The present invention introduces deep learning algorithms and image processing technology to provide an efficient and automated cigarette stain detection system that can detect stains on the surface of cigarettes in real time and generate corresponding annotated images, greatly improving the detection accuracy and efficiency, and is suitable for the quality control process of modern tobacco production lines.
Claims
1. A system for detecting cigarette stains, characterized in that: It includes an image acquisition module for taking images of cigarettes; Contains a deep learning processing module for processing the captured images; The deep learning processing module analyzes the image through a pre-trained stain detection model to identify stains on the surface of the cigarette; It includes a stain location module, which uses the analysis results to identify the location of the stain and generates an image output with the stain location marked.
2. The system according to claim 1, characterized in that The deep learning processing module uses a convolutional neural network (CNN) model to perform image processing and stain recognition.
3. The system according to claim 1 or 2, characterized in that: The stain detection model is trained by including a large number of annotated cigarette stain datasets.
4. The system according to any one of claims 1 to 3, characterized in that: The image acquisition module includes a high-resolution camera to ensure the accuracy of stain detection.
5. The system according to any one of claims 1 to 4, characterized in that: The stain location module can generate a rectangular bounding box to accurately mark the stain location.
6. The system according to any one of claims 1 to 5, characterized in that: The system supports real-time detection and is applied to the quality detection link in the cigarette production process through an assembly line method.
7. The system according to any one of claims 1 to 6, characterized in that: The system has a data storage function and can save images and stain location information during the detection process for subsequent analysis.