A cigarette burning rate determination method, device, equipment and storage medium
By segmenting cigarette combustion images using deep learning technology and image segmentation models, the problem of complexity and limited accuracy in traditional cigarette combustion rate measurement methods is solved, achieving efficient and accurate measurement of cigarette combustion rate.
Patent Information
- Application Number
- CN202411989795.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-30
- Publication Date
- 2026-06-30
AI Technical Summary
Traditional methods for measuring cigarette burn rate involve complex equipment and have limited accuracy, making it difficult to efficiently and accurately assess cigarette quality and combustion characteristics.
Deep learning technology is used to segment images of burned cigarettes using an image segmentation model. The burned and unburned areas of the cigarettes are determined by a trained U-Net model. The burned and unburned areas are calculated by combining camera calibration parameters, and then the combustion rate is calculated.
It achieves efficient and accurate determination of cigarette combustion rate, avoids interference with the combustion process, and improves measurement efficiency and accuracy.
Smart Images
Figure CN122312474A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of cigarette combustion characteristic measurement technology, specifically to a method, apparatus, equipment and storage medium for determining the combustion rate of cigarettes. Background Technology
[0002] Accurate determination of cigarette combustion rate is of great significance for assessing cigarette quality, combustion characteristics, and environmental impact.
[0003] However, traditional methods for determining the burning rate of cigarettes often require complex equipment and cumbersome procedures, and have limited accuracy. Summary of the Invention
[0004] In view of this, the embodiments of this application aim to provide a method, apparatus, device and storage medium for measuring the burning rate of cigarettes, which can improve the efficiency and accuracy of measuring the burning rate of cigarettes.
[0005] According to a first aspect of the embodiments of this application, a method for determining the burning rate of a cigarette is provided, the method comprising:
[0006] The burning image of the cigarette to be measured is input into the trained image segmentation model, and the segmentation result corresponding to the burning image of the cigarette to be measured is output; the burning image of the cigarette to be measured is obtained by preprocessing the image corresponding to the burning cigarette; the image segmentation model is obtained by training based on the burning image samples of the cigarette and the semantic labels of the corresponding pixels.
[0007] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the combustion rate of the cigarette is determined.
[0008] Optionally, determining the cigarette burning rate based on the segmentation results corresponding to the cigarette burning image to be measured includes:
[0009] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined;
[0010] The combustion rate of a cigarette is determined based on its burning and unburned areas.
[0011] Optionally, determining the burned area and unburned area of the cigarette based on the segmentation results corresponding to the burning image of the cigarette to be measured includes:
[0012] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined;
[0013] The burning area and unburned area of the cigarette are determined based on the number of pixels corresponding to the burning and unburned areas of the cigarette and the camera calibration parameters corresponding to the burning image of the cigarette to be measured.
[0014] Optionally, based on the segmentation results corresponding to the burning image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined, including:
[0015] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burned area, the unburned area, and the overlapping area of the burned area and the unburned area of the cigarette are determined.
[0016] Based on the number of pixels corresponding to the burning and unburned areas of the cigarette and the camera calibration parameters corresponding to the burning image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined, including:
[0017] Based on the number of pixels corresponding to the burned area, the unburned area and the overlapping area of the cigarette, and the camera calibration parameters corresponding to the burning image of the cigarette to be measured, the burned area, the unburned area and the overlapping area are determined.
[0018] The sum of the burned area and the overlapping area is determined as the burning area.
[0019] Optionally, the cigarette burning image samples include subsets of cigarette burning image samples corresponding to multiple brands of cigarettes. For any brand of cigarettes, the steps for obtaining the corresponding subset of cigarette burning image samples include:
[0020] Multiple combustion experiments were conducted on the current brand of cigarettes, and a set of cigarette combustion images corresponding to different combinations of combustion states, lighting conditions, and shooting angles was obtained.
[0021] Data augmentation and preprocessing are performed on each cigarette burning image in the cigarette burning image set to obtain a subset of cigarette burning image samples corresponding to the current brand of cigarettes.
[0022] Optionally, in the process of obtaining the image segmentation model, based on the subset of cigarette burning image samples from the multiple brands of cigarettes, a subset of cigarette burning image samples from one brand of cigarettes is used as a validation set, and a subset of cigarette burning image samples from other brands of cigarettes is used as a training set to train the image segmentation model to obtain a new cigarette burning image segmentation model, so as to determine the cigarette burning rate of the one brand of cigarettes in the validation set based on the new cigarette burning image segmentation model.
[0023] Optionally, the data augmentation includes random rotation, random flipping, random scaling, and random cropping of the burning images of each cigarette, and the preprocessing includes image denoising, contrast enhancement, and min-max normalization.
[0024] Optionally, the image segmentation model is the U-Net model.
[0025] Optionally, the U-Net model uses the Dice loss function during training and combines it with the Adam optimization algorithm to automatically adjust the model parameters.
[0026] According to a second aspect of the embodiments of this application, a cigarette burning rate measuring device is provided, comprising:
[0027] The image segmentation module is used to input the burning image of the cigarette to be measured into the trained image segmentation model and output the segmentation result corresponding to the burning image of the cigarette to be measured; the image segmentation model is obtained by training based on the burning image samples of the cigarette and the semantic labels of the corresponding pixels.
[0028] The cigarette burning rate determination module is used to determine the cigarette burning rate based on the segmentation results corresponding to the burning image of the cigarette to be measured.
[0029] According to a third aspect of the present application, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; the processor being configured to perform the method described in any of the above embodiments.
[0030] According to a fourth aspect of the present application, a computer-readable storage medium is provided, the storage medium storing a computer program for performing the methods described in any of the above embodiments.
[0031] This application provides a method, apparatus, device, and storage medium for determining the burning rate of a cigarette. The scheme includes: inputting a burning image of a cigarette to be measured into a trained image segmentation model, and outputting a segmentation result corresponding to the burning image; the burning image of the cigarette to be measured is obtained by preprocessing an image corresponding to a burned cigarette; the image segmentation model is obtained by training based on burning image samples and the semantic labels of the corresponding pixels; and determining the burning rate of the cigarette based on the segmentation result corresponding to the burning image. This method enables analysis of the burning rate of a cigarette based on the burning image, avoiding interference with the cigarette burning process. Simultaneously, it uses a deep learning model for accurate and efficient segmentation of the burning area, ensuring the efficiency and accuracy of the burning rate measurement. Attached Figure Description
[0032] Figure 1 The diagram shown is a flowchart of a cigarette burning rate determination method provided in one embodiment of this application.
[0033] Figure 2 The image shown is a schematic diagram of a cigarette burning image sample and the corresponding pixel semantic labels provided in an embodiment of this application.
[0034] Figure 3The diagram shown is a block diagram of a cigarette burning rate measuring device provided in one embodiment of this application.
[0035] Figure 4 The diagram shown is a structural block diagram of an electronic device provided in one embodiment of this application. Detailed Implementation
[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0037] Exemplary methods
[0038] Figure 1 This is a schematic flowchart of a method for determining the burning rate of a cigarette according to an embodiment of this application. Figure 1 As shown, the method includes the following:
[0039] Step 101: Input the burning image of the cigarette to be measured into the trained image segmentation model, and output the segmentation result corresponding to the burning image of the cigarette to be measured; the burning image of the cigarette to be measured is obtained by preprocessing the image corresponding to the burning cigarette; the image segmentation model is obtained by training based on the burning image samples of the cigarette and the semantic labels of the corresponding pixels.
[0040] Specifically, as mentioned above, traditional methods for determining cigarette burn rate often require complex equipment and cumbersome procedures, and have limited accuracy. This application, through research, has found that with the development of deep learning technology and its excellent performance in image segmentation, using deep convolutional neural network models for image recognition provides a new approach to determining cigarette burn rate. Based on this, this application proposes a method for determining cigarette burn rate, which uses a deep convolutional neural network model to accurately and efficiently segment images of burned cigarettes for accurate burn rate determination.
[0041] More specifically, in this embodiment of the application, images of cigarettes burning in different brands and under different burning conditions are collected in advance, and the samples are labeled to obtain semantic labels for the corresponding pixels. Then, the image segmentation model is trained based on the cigarette burning image samples and the corresponding pixel semantic labels to obtain a trained image segmentation model. Figure 2The diagram illustrates a sample of a burning cigarette image and its corresponding pixel semantic labels, provided in an embodiment of this application. The left side shows the burning cigarette image sample, and the right side shows the annotation results. A trained image segmentation model can accurately segment a burning cigarette image, producing a binary image where the burning area is represented by a value of 1 and the unburned area by a value of 0. Based on this, this embodiment can quickly determine the burning and unburned areas in the burning cigarette image to be measured using the image segmentation model, thereby ensuring accurate calculation of the cigarette's combustion rate. It is understood that the burning cigarette image to be measured is obtained by preprocessing the image corresponding to the burned cigarette. The image corresponding to the burned cigarette can be acquired using a high-resolution camera. Specifically, the camera's focal length, aperture, and ISO can be adjusted according to the environment to ensure that the captured image clearly presents the state of the burned cigarette, providing high-quality image data for subsequent analysis and processing, such as accurately capturing the remaining length of the cigarette, the distribution of ash, and the texture features of the burned surface. The purpose of the preprocessing is to improve image quality, reduce environmental and camera-induced light sensitivity errors, highlight the characteristics of the cigarette burning area, and thus ensure the efficiency and accuracy of subsequent image segmentation.
[0042] Step 102: Determine the cigarette combustion rate based on the segmentation results corresponding to the cigarette combustion image to be measured.
[0043] Specifically, based on the foregoing, after obtaining the segmentation results corresponding to the burning image of the cigarette to be measured, the burning area and the unburned area can be determined, thereby quickly determining the cigarette burning rate.
[0044] In this embodiment, the burning image of the cigarette to be measured is input into a trained image segmentation model, which outputs the segmentation result corresponding to the burning image of the cigarette to be measured. The burning image of the cigarette to be measured is obtained by preprocessing the image corresponding to the burned cigarette. The image segmentation model is obtained by training based on cigarette burning image samples and the semantic labels of the corresponding pixels. Based on the segmentation result corresponding to the burning image of the cigarette to be measured, the burning rate of the cigarette is determined. This allows for the analysis of the burning rate of the cigarette based on the burning image, avoiding interference with the cigarette burning process. At the same time, the deep learning model performs accurate and efficient segmentation of the burning area, ensuring the efficiency and accuracy of the cigarette burning rate measurement.
[0045] based on Figure 1 In addition to the method described in the embodiments of this specification, some specific implementation schemes of the method are also provided, which will be described below.
[0046] Optionally, determining the cigarette burning rate based on the segmentation results corresponding to the cigarette burning image to be measured includes:
[0047] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined;
[0048] The combustion rate of a cigarette is determined based on its burning and unburned areas.
[0049] Specifically, based on the foregoing, after determining the segmentation result corresponding to the burning image of the cigarette to be measured, the burning area and unburned area of the cigarette can be determined based on the segmentation result. Furthermore, the total area of the cigarette can be determined based on the burning area and unburned area, and the cigarette combustion rate is the quotient of the burning area and the total area.
[0050] Optionally, determining the burned area and unburned area of the cigarette based on the segmentation results corresponding to the burning image of the cigarette to be measured includes:
[0051] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined;
[0052] The burning area and unburned area of the cigarette are determined based on the number of pixels corresponding to the burning and unburned areas of the cigarette and the camera calibration parameters corresponding to the burning image of the cigarette to be measured.
[0053] Specifically, it can be understood that the segmentation results corresponding to the cigarette combustion image to be measured determine the burning and unburned areas of the cigarette in the burning image. However, to calculate the cigarette combustion rate, it is necessary to determine the actual burning and unburned areas of the cigarette. Based on this, this embodiment further determines the burning and unburned areas of the cigarette based on the number of pixels corresponding to the burning and unburned areas and the camera calibration parameters corresponding to the cigarette combustion image to be measured. The camera calibration parameters corresponding to the cigarette combustion image to be measured are the calibration parameters of the high-resolution camera that acquired the cigarette combustion image to be measured. For the method of calculating the actual area based on the number of pixels and camera calibration parameters, any existing feasible method can be used as needed, and this embodiment does not specifically limit it here.
[0054] Optionally, during the actual combustion process of a cigarette, there will be a burned area, a burning area, and an unburned area. Therefore, determining the burning area and unburned area of a cigarette according to the above-mentioned area division method is more granular and more accurate.
[0055] For example, when determining the burning and unburned areas of a cigarette based on the segmentation results corresponding to the burning image of the cigarette to be measured, the burned area, unburned area, and overlapping area of the burned and unburned areas of the cigarette can be determined based on the segmentation results corresponding to the burning image of the cigarette to be measured.
[0056] The burned area is the area that has been completely burned, the unburned area is the area that has not yet been burned, and the overlapping area of the burned and unburned areas is the area that is currently burning.
[0057] The size of the overlapping area between the burned and unburned areas can be flexibly adjusted by setting an overlap threshold. For example, the area formed by setting a width on both sides of the boundary between the burned and unburned areas can be used as the overlapping area.
[0058] Based on the determination of the burned area, unburned area, and overlapping area, when determining the burned area and unburned area of a cigarette, the burned area, unburned area, and overlapping area are determined based on the number of pixels corresponding to each of the burned area, unburned area, and overlapping area, as well as the camera calibration parameters corresponding to the burning image of the cigarette to be measured. For example, by converting the number of pixels corresponding to each of the burned area, unburned area, and overlapping area into actual physical area values using the camera calibration parameters, the burned area, unburned area, and overlapping area can be obtained.
[0059] Then, the sum of the burned area and the overlapping area is determined as the burning area.
[0060] Optionally, the cigarette burning image samples include subsets of cigarette burning image samples corresponding to multiple brands of cigarettes. For any brand of cigarettes, the steps for obtaining the corresponding subset of cigarette burning image samples include:
[0061] Multiple combustion experiments were conducted on the current brand of cigarettes, and a set of cigarette combustion images corresponding to different combinations of combustion states, lighting conditions, and shooting angles was obtained.
[0062] Data augmentation and preprocessing are performed on each cigarette burning image in the cigarette burning image set to obtain a subset of cigarette burning image samples corresponding to the current brand of cigarettes.
[0063] Specifically, to ensure the diversity and representativeness of the data, the cigarette burning image samples in this application include subsets of cigarette burning image samples corresponding to multiple brands of cigarettes. For any brand of cigarette, multiple burning experiments are conducted to obtain a set of cigarette burning images corresponding to different combinations of burning states, lighting conditions, and shooting angles. Furthermore, to further increase the quantity and diversity of training data, this application performs data augmentation operations on each cigarette burning image in the set to improve the generalization ability of the image segmentation model. Specific data augmentation operations include, but are not limited to, random rotation, random flipping, random scaling, and random cropping. Simultaneously, this application also performs the aforementioned preprocessing on each cigarette burning image in the set to ensure image quality.
[0064] Optionally, based on the aforementioned subsets of cigarette burning images corresponding to multiple brands of cigarettes, during the training of the image segmentation model, a subset of cigarette burning images from one brand of cigarettes is used as the validation set, and subsets of cigarette burning images from other brands of cigarettes are used as the training set. The image segmentation model is then trained to obtain a new cigarette burning image segmentation model, and the burning rate of the cigarettes of the specified brand in the validation set is determined based on the new cigarette burning image segmentation model.
[0065] That is, the cigarette burning image segmentation model is trained using a subset of cigarette burning image samples from other brands in the training set. After training, the newly trained cigarette burning image segmentation model is used to determine the burning rate of a specific brand of cigarettes in the validation set. Specifically, the burning images of a specific brand of cigarettes from the validation set are input into the new cigarette burning image segmentation model, which then segments the images to obtain the segmentation result. The burning rate is then determined based on this segmentation result.
[0066] The model training method described above enables the model to accurately measure the combustion rate of cigarettes of any brand, thus improving its applicability.
[0067] Optionally, the preprocessing includes image denoising, contrast enhancement, and min-max normalization.
[0068] Specifically, image denoising and min-max normalization aim to improve image quality, while contrast enhancement reduces environmental and camera-induced light sensitivity errors, highlighting the characteristics of the cigarette's burning area. These preprocessing operations maximize the accuracy of subsequent image segmentation model processing. Contrast enhancement can be implemented using a local histogram equalization algorithm.
[0069] Optionally, the image segmentation model is the U-Net model.
[0070] Specifically, the U-Net model includes an encoder, a decoder, and skip connections. By utilizing a symmetrical U-shaped decoder and encoder structure, and passing features across paths through skip connections, it can effectively extract and utilize multi-level features of images, improving the accuracy and robustness of segmentation.
[0071] The encoder consists of multiple convolutional layers repeated twice in the same stage and four downsampling layers to extract high-level features from the image. Each convolutional layer uses an appropriately sized kernel, such as a 3x3 or 5x5 kernel, to capture local features of the image. The kernel size can be adjusted according to the image size to obtain the optimal receptive field at the same resolution. The downsampling layers use methods such as max pooling to progressively reduce the image resolution, thereby progressively expanding the receptive field and extracting more global features.
[0072] The decoder consists of four upsampling layers and multiple convolutional layers that are repeated twice in the same stage. It is used to restore the image resolution and perform accurate segmentation. The upsampling layers employ methods such as deconvolution to progressively increase the image resolution, restoring low-resolution feature maps to the same size as the input image. Each convolutional layer further refines the segmentation results, improving segmentation accuracy.
[0073] Skip connections, established between the encoder and decoder, fuse features from different stages. Skip connections effectively transfer low-level detail information and high-level semantic information, improving the model's image segmentation capabilities.
[0074] Optionally, the U-Net model uses the Dice loss function during training and combines it with the Adam optimization algorithm to automatically adjust the model parameters.
[0075] Specifically, in the U-Net model training phase of this application embodiment, the model parameters, such as convolutional kernel size, number of layers, and learning rate, are first set, and then the prepared training data is used to train the model. Specifically, this application embodiment divides the aforementioned cigarette burning image samples and their corresponding pixel semantic labels into a training set, an internal validation set, and a test set. The training set is used to train the model, helping it learn the mapping relationship between input data and output labels. Through backpropagation and optimization using the Dice loss function, the model automatically adjusts the feature extraction parameters to fit the training data and output accurate segmentation results. The purpose of model training is to ensure that the model can effectively capture important features in the image, improving its performance in cigarette burning segmentation tasks. The formula for calculating the Dice loss function is:
[0076]
[0077] Where N represents the total number of pixels in the sample, y i This represents the i-th pixel value of the actual label. This represents the i-th probability value of the predicted label. The smaller the value of the Dice loss function, the closer the predicted result is to the true result.
[0078] Meanwhile, this embodiment employs the Adam optimization algorithm to continuously adjust the model's parameters, making the model's output as close as possible to the labeled true values. During training, the model is evaluated using an internal validation set every 1-3 epochs to monitor its performance. Evaluation metrics include Dice coefficient, accuracy, recall, and intersection-over-union ratio. Based on the evaluation results, the model's structure and parameters are adjusted, or other optimization strategies are adopted, such as adjusting the learning rate or increasing the degree of data augmentation, to further improve the model's performance. More specifically, this embodiment preferably sets the learning rate to 1e-5 to 1e-3 and decreases slowly with the training process, the batch size to 12-16, and the number of training epochs to 100-50000 to ensure that the model can converge quickly and achieve high accuracy.
[0079] The test set is used to evaluate the performance and generalization ability of the image segmentation model. The data in the test set is independent of the training set, meaning that the test set contains data that the model has never seen before, thus simulating real-world application scenarios.
[0080] Exemplary device
[0081] The apparatus embodiments of this application can be used to execute the method embodiments of this application. For details not disclosed in the apparatus embodiments of this application, please refer to the method embodiments of this application.
[0082] Figure 3 The diagram shown is a block diagram of a cigarette burning rate measuring device according to an embodiment of this application. Figure 3 As shown, the device 200 includes:
[0083] The image segmentation module 201 is used to input the cigarette burning image to be measured into the trained image segmentation model and output the segmentation result corresponding to the cigarette burning image to be measured; the image segmentation model is obtained by training based on cigarette burning image samples and the semantic labels of the corresponding pixels.
[0084] The cigarette burning rate determination module 202 is used to determine the cigarette burning rate based on the segmentation results corresponding to the burning image of the cigarette to be measured.
[0085] Optionally, determining the cigarette burning rate based on the segmentation results corresponding to the cigarette burning image to be measured includes:
[0086] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined;
[0087] The combustion rate of a cigarette is determined based on its burning and unburned areas.
[0088] Optionally, determining the burned area and unburned area of the cigarette based on the segmentation results corresponding to the burning image of the cigarette to be measured includes:
[0089] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined;
[0090] The burning area and unburned area of the cigarette are determined based on the number of pixels corresponding to the burning and unburned areas of the cigarette and the camera calibration parameters corresponding to the burning image of the cigarette to be measured.
[0091] Optionally, based on the segmentation results corresponding to the burning image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined, including:
[0092] Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burned area, the unburned area, and the overlapping area of the burned area and the unburned area of the cigarette are determined.
[0093] Based on the number of pixels corresponding to the burning and unburned areas of the cigarette and the camera calibration parameters corresponding to the burning image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined, including:
[0094] Based on the number of pixels corresponding to the burned area, the unburned area and the overlapping area of the cigarette, and the camera calibration parameters corresponding to the burning image of the cigarette to be measured, the burned area, the unburned area and the overlapping area are determined.
[0095] The sum of the burned area and the overlapping area is determined as the burning area.
[0096] Optionally, the cigarette burning image sample includes subsets of cigarette burning image samples corresponding to multiple brands of cigarettes. The device further includes a cigarette burning image sample subset acquisition module, which is used to perform the following operations on any brand of cigarettes:
[0097] Multiple combustion experiments were conducted on the current brand of cigarettes, and a set of cigarette combustion images corresponding to different combinations of combustion states, lighting conditions, and shooting angles was obtained.
[0098] Data augmentation and preprocessing are performed on each cigarette burning image in the cigarette burning image set to obtain a subset of cigarette burning image samples corresponding to the current brand of cigarettes.
[0099] Optionally, in the process of obtaining the image segmentation model, based on the subset of cigarette burning image samples from the multiple brands of cigarettes, a subset of cigarette burning image samples from one brand of cigarettes is used as a validation set, and a subset of cigarette burning image samples from other brands of cigarettes is used as a training set to train the image segmentation model to obtain a new cigarette burning image segmentation model, so as to determine the cigarette burning rate of the one brand of cigarettes in the validation set based on the new cigarette burning image segmentation model.
[0100] Optionally, the data augmentation includes random rotation, random flipping, random scaling, and random cropping of the burning images of each cigarette, and the preprocessing includes image denoising, contrast enhancement, and min-max normalization.
[0101] Optionally, the image segmentation model is the U-Net model.
[0102] Optionally, the U-Net model uses the Dice loss function during training and combines it with the Adam optimization algorithm to automatically adjust the model parameters.
[0103] Exemplary electronic devices
[0104] Below, for reference Figure 4 This describes an electronic device according to embodiments of the present application. Figure 4 A block diagram of an electronic device according to an embodiment of this application is illustrated.
[0105] like Figure 4 As shown, the electronic device 300 includes one or more processors 301 and memory 302.
[0106] The processor 301 may be a central processing unit (CPU) or other form of processing unit with data processing and / or instruction execution capabilities, and may control other components in the electronic device 300 to perform desired functions.
[0107] The memory 302 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 301 may execute the program instructions to implement the cigarette burning rate determination methods of the various embodiments of this application described above and / or other desired functions. Various contents, such as category correspondence, may also be stored in the computer-readable storage medium.
[0108] In one example, the electronic device 300 may also include an input device 303 and an output device 304, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).
[0109] In addition, the input device 303 may also include, for example, a keyboard, a mouse, etc. The output device 304 can output various information to the outside. The output device 304 may include, for example, a monitor, a speaker, a printer, and a communication network and its connected remote output devices, etc.
[0110] Of course, for the sake of simplicity, Figure 4 Only some of the components of the electronic device 300 relevant to this application are shown in this illustration; components such as buses, input / output interfaces, etc., are omitted. In addition, the electronic device 300 may include any other suitable components depending on the specific application.
[0111] Exemplary computer program products and computer-readable storage media
[0112] In addition to the methods and apparatus described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps in the cigarette burning rate determination methods according to various embodiments of this application described in the "Exemplary Methods" section of this specification.
[0113] The computer program product can be written in any combination of one or more programming languages to perform the operations of the embodiments of this application. The programming languages include object-oriented programming languages such as Java and C++, as well as conventional procedural programming languages such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.
[0114] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps in the cigarette burning rate determination methods according to various embodiments of this application described in the "Exemplary Methods" section of this specification.
[0115] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may, for example, include, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatuses, or devices, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: electrical connections having one or more wires, portable disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.
[0116] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.
[0117] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.
[0118] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.
[0119] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.
[0120] It should be understood that the qualifiers “first,” “second,” “third,” “fourth,” “fifth,” and “sixth” used in the description of the embodiments of this application are only used to more clearly illustrate the technical solutions and are not intended to limit the scope of protection of this application.
[0121] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.
Claims
1. A method for determining the burning rate of a cigarette, characterized in that, The method includes: The burning image of the cigarette to be measured is input into the trained image segmentation model, and the segmentation result corresponding to the burning image of the cigarette to be measured is output; the burning image of the cigarette to be measured is obtained by preprocessing the image corresponding to the burning cigarette; the image segmentation model is obtained by training based on the burning image samples of the cigarette and the semantic labels of the corresponding pixels. Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the combustion rate of the cigarette is determined.
2. The method according to claim 1, characterized in that, The step of determining the cigarette combustion rate based on the segmentation results corresponding to the cigarette combustion image to be measured includes: Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined; The combustion rate of a cigarette is determined based on its burning and unburned areas.
3. The method according to claim 2, characterized in that, The step of determining the burned area and unburned area of the cigarette based on the segmentation results corresponding to the burning image of the cigarette to be measured includes: Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined; The burning area and unburned area of the cigarette are determined based on the number of pixels corresponding to the burning and unburned areas of the cigarette and the camera calibration parameters corresponding to the burning image of the cigarette to be measured.
4. The method according to claim 3, characterized in that, Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined, including: Based on the segmentation results corresponding to the combustion image of the cigarette to be measured, the burned area, the unburned area, and the overlapping area of the burned area and the unburned area of the cigarette are determined. Based on the number of pixels corresponding to the burning and unburned areas of the cigarette and the camera calibration parameters corresponding to the burning image of the cigarette to be measured, the burning area and unburned area of the cigarette are determined, including: Based on the number of pixels corresponding to the burned area, the unburned area and the overlapping area of the cigarette, and the camera calibration parameters corresponding to the burning image of the cigarette to be measured, the burned area, the unburned area and the overlapping area are determined. The sum of the burned area and the overlapping area is determined as the burning area.
5. The method according to claim 1, characterized in that, The cigarette burning image samples include subsets of cigarette burning image samples corresponding to multiple brands of cigarettes. For any brand of cigarettes, the steps for obtaining the corresponding subset of cigarette burning image samples include: Multiple combustion experiments were conducted on the current brand of cigarettes, and a set of cigarette combustion images corresponding to different combinations of combustion states, lighting conditions, and shooting angles was obtained. Data augmentation and preprocessing are performed on each cigarette burning image in the cigarette burning image set to obtain a subset of cigarette burning image samples corresponding to the current brand of cigarettes.
6. The method according to claim 5, characterized in that, In the process of obtaining the image segmentation model, based on the subset of cigarette burning image samples of the multiple brands of cigarettes, a subset of cigarette burning image samples of one brand of cigarettes is used as the validation set, and a subset of cigarette burning image samples of other brands of cigarettes is used as the training set. The image segmentation model is trained to obtain a new cigarette burning image segmentation model, so as to determine the cigarette burning rate of the one brand of cigarettes in the validation set based on the new cigarette burning image segmentation model.
7. The method according to claim 5, characterized in that, The data augmentation includes random rotation, random flipping, random scaling, and random cropping of the burning images of each cigarette. The preprocessing includes image denoising, contrast enhancement, and min-max normalization.
8. The method according to claim 1, characterized in that, The image segmentation model is the U-Net model.
9. The method according to claim 8, characterized in that, The U-Net model uses the Dice loss function during training and combines it with the Adam optimization algorithm to automatically adjust the model parameters.
10. A device for measuring the burning rate of cigarettes, characterized in that, include: The image segmentation module is used to input the burning image of the cigarette to be measured into the trained image segmentation model and output the segmentation result corresponding to the burning image of the cigarette to be measured; the image segmentation model is obtained by training based on the burning image samples of the cigarette and the semantic labels of the corresponding pixels. The cigarette burning rate determination module is used to determine the cigarette burning rate based on the segmentation results corresponding to the burning image of the cigarette to be measured.
11. An electronic device, comprising: processor; Memory used to store the processor's executable instructions; The processor is configured to perform the method according to any one of claims 1 to 9.
12. A computer-readable storage medium storing computer program instructions that, when executed by a processor, cause the processor to perform the method as described in any one of claims 1 to 9.