Agilawood koji production quality control method based on image recognition technology

Abstract

The invention relates to an agilawood koji production quality control method based on an image recognition technology, and the method carries out the feature training learning of the texture features of an agilawood koji slice image through a big data technology, constructs a quality inspection recognition model, and is used for recognizing the quality grade of the agilawood koji slice image. The AI model can be utilized to identify the texture features of the agilawood yeast slice image and output the corresponding quality grade, the AI technology is utilized to perform image feature identification, and the model outputs the corresponding agilawood yeast processing quality grade. Therefore, the model can be combined to carry out image identification and fingerprint identification on massive agilawood starter slice images, and the identification efficiency and accuracy of the agilawood starter are greatly improved. By adopting the scheme, the entry threshold and requirements of identification can be reduced, mass quality detection can be realized, and the processing quality delivery standard of the agilawood koji is improved.

Description

Technical Field

[0001] The present disclosure relates to the field of image recognition technology, and in particular to a method, system and electronic equipment for controlling the quality of agarwood production based on image recognition technology. Background Art

[0002] The main effects of agarwood music are as follows:

[0003] Dispel wind and relieve cold: Agarwood contains a variety of warm medicinal materials, which can warm the body's yang energy, promote blood circulation, and have a good effect on cold diseases such as colds, arthritis, etc.

[0004] Soothing the mind and calming the spirit: Its volatile oil components can calm the nerves, soothe emotions, help treat insomnia, anxiety, depression and other symptoms, and help restore a calm state.

[0005] Anti-inflammatory and antibacterial: Quercetin, quercetin and other ingredients in agarwood have antibacterial and anti-inflammatory effects, which can effectively inhibit the growth of pathogenic microorganisms and reduce inflammatory reactions, such as oral ulcers, sores, swelling and pain.

[0006] Improve gastrointestinal function: It can enhance gastric juice secretion, promote digestion, improve abdominal distension and constipation, regulate intestinal peristalsis, and relieve gastrointestinal discomfort.

[0007] The production and processing quality of agarwood koji has a significant impact on its efficacy, so it is particularly important to control the production and processing quality of agarwood koji. Among them: the production of agarwood koji requires the use of high-quality agarwood and other medicinal materials as raw materials to ensure the purity and efficacy of the raw materials; it must be processed strictly in accordance with traditional processing methods, including the grinding, mixing, paste making, molding and drying of medicinal materials. Each step must be precisely controlled to ensure the quality of agarwood koji; modern analytical techniques such as high performance liquid chromatography are used to test the quality of agarwood koji, and by comparing fingerprint spectra, its quality is judged to be qualified to ensure the consistency and stability of the product.

[0008] In the process of quality inspection of agarwood, the traditional quality inspection scheme requires manual identification of the fingerprint of agarwood (usually slices), which will have the following technical defects:

[0009] First, manual fingerprint identification requires a wealth of industry experience and a high level of fingerprint identification, so the technology is difficult and the entry requirements for identification personnel are high;

[0010] Second, faced with the fingerprint identification of massive slices, manual work can only be done by random sampling. This method is not only inefficient, but also unable to achieve large-scale identification and inspection. Therefore, the production quality of agarwood wine cannot be well controlled. Summary of the invention

[0011] In order to solve the above problems, the present application proposes a method, system and electronic equipment for controlling the production quality of agarwood music based on image recognition technology.

[0012] On the one hand, the present application proposes a method for controlling the quality of agarwood production based on image recognition technology, comprising the following steps:

[0013] S1, collecting images of agarwood slices;

[0014] S2, inputting the agarwood slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood slice image and outputs the corresponding quality grade;

[0015] S3, judging whether the quality level of the current agarwood slices meets the standard:

[0016] If the standard is met, the next agarwood slice will be quality inspected;

[0017] Otherwise, an alarm signal is issued.

[0018] As an optional implementation scheme of the present application, optionally, S1, collecting an image of agarwood song slice, comprises:

[0019] Batch collect the agarwood slice images of a number of agarwood slices and pre-process them;

[0020] Slicing the agarwood song slice image to obtain image slices corresponding to a number of agarwood song slices;

[0021] Check the position correspondence between the image slice and the corresponding agarwood song slice, and record the position information of the image slice.

[0022] As an optional implementation scheme of the present application, optionally, the method for generating the quality inspection recognition model includes:

[0023] Collect and pre-process several agarwood slice images;

[0024] Traversing each of the agarwood slice images, and using CNN to extract texture features in the agarwood slice images, including glossiness, texture spatial distribution, and texture morphology;

[0025] Performing feature annotation on the texture features, wherein the annotation information includes a quality grade determined based on the glossiness, texture spatial distribution, and texture morphology;

[0026] Constructing a feature set based on the labeled texture features;

[0027] Dividing the feature set into a training set and a validation set according to a preset ratio;

[0028] Input the training set into a preset CNN model to perform feature training and learning to construct the quality inspection recognition model;

[0029] The recognition performance of the quality inspection recognition model is verified using the verification set:

[0030] If the verification is passed, the quality inspection and recognition model is deployed on the server and used for texture feature quality recognition of the agarwood song slice image;

[0031] If the verification fails, repeat the above steps to rebuild the quality inspection recognition model.

[0032] As an optional implementation scheme of the present application, optionally, the annotation information further includes:

[0033] The unit price is determined based on the glossiness, texture spatial distribution, and texture morphology.

[0034] As an optional implementation scheme of the present application, optionally, S2, inputting the agarwood song slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood song slice image and outputs the corresponding quality level, including:

[0035] Inputting the agarwood song slice image into the quality inspection and recognition model;

[0036] Identify the texture features in the agarwood song slice image through the quality inspection recognition model, and predict and output the quality level corresponding to the texture features;

[0037] The quality grade is bound to the corresponding agarwood song slice image and written into the corresponding image table to obtain an agarwood song quality identification table.

[0038] As an optional implementation scheme of the present application, optionally, S2, inputting the agarwood song slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood song slice image and outputs the corresponding quality level, including:

[0039] Inputting the agarwood song slice image into the quality inspection and recognition model;

[0040] Identify the texture features in the agarwood slice image through the quality inspection recognition model, and predict and output the unit price corresponding to the texture features;

[0041] The unit price is bound to the corresponding agarwood song slice image and written into the corresponding image table to obtain an agarwood song unit price prediction table.

[0042] On the other hand, the present application proposes a quality control system for agarwood production based on image recognition technology, comprising:

[0043] An image acquisition module is used to collect images of agarwood slices and input them into the agarwood quality inspection module;

[0044] Agarwood quality inspection module, used for inputting the agarwood slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood slice image and outputs the corresponding quality grade;

[0045] The quality control module is used to determine whether the quality level of the current agarwood slices meets the standards:

[0046] If the standard is met, the next agarwood slice will be quality inspected;

[0047] Otherwise, an alarm signal is issued;

[0048] An alarm module, used for responding to the alarm signal;

[0049] The image acquisition module, agarwood song quality inspection module, quality control module and alarm module are communicatively connected in sequence.

[0050] On the other hand, the present application also proposes an electronic device, comprising:

[0051] processor;

[0052] a memory for storing processor-executable instructions;

[0053] Among them, the processor is configured to implement the agarwood song production quality control method based on image recognition technology when executing the executable instructions.

[0054] Technical effects of the present invention:

[0055] The present invention uses big data technology to perform feature training and learning on the texture features of agarwood song slice images, construct a quality inspection and recognition model and apply it to identify the quality level of agarwood song slice images. The AI ​​model can be used to identify the texture features of agarwood song slice images and output the corresponding quality level, and the AI ​​technology can be used to perform image feature recognition, so that the model outputs the corresponding agarwood song processing quality level. Therefore, the model can be combined to perform image recognition and fingerprint identification on massive agarwood song slice images, greatly improving the efficiency and accuracy of agarwood song identification. The adoption of this solution can lower the entry threshold and requirements for identification, realize large-scale quality inspection, and improve the processing quality outbound standards of agarwood song.

[0056] Further features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate exemplary embodiments, features, and aspects of the disclosure and, together with the description, serve to explain the principles of the disclosure.

[0058] Figure 1 It is shown as a schematic diagram of the implementation process of the present invention;

[0059] Figure 2 Shown is a flow chart of the quality inspection recognition model training of the present invention;

[0060] Figure 3 Shown is a schematic diagram of the system composition structure of the present invention;

[0061] Figure 4 The application diagram of the electronic device of the present invention is shown as follows DETAILED DESCRIPTION

[0062] Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numerals in the accompanying drawings represent elements with the same or similar functions. Although various aspects of the embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless otherwise specified.

[0063] The word “exemplary” is used exclusively herein to mean “serving as an example, example, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments.

[0064] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the main purpose of the present disclosure.

[0065] Example 1

[0066] like Figure 1 As shown, on the one hand, the present application proposes a method for controlling the quality of agarwood production based on image recognition technology, comprising the following steps:

[0067] S1, collecting images of agarwood slices;

[0068] S2, inputting the agarwood slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood slice image and outputs the corresponding quality grade;

[0069] S3, judging whether the quality level of the current agarwood slices meets the standard:

[0070] If the standard is met, the next agarwood slice will be quality inspected;

[0071] Otherwise, an alarm signal is issued.

[0072] The present invention uses big data technology to perform feature training and learning on the texture features of agarwood slice images, construct a quality inspection and recognition model and apply it to identify the quality level of agarwood slice images. The AI ​​model can be used to identify the texture features of agarwood slice images and output the corresponding quality level, and the AI ​​technology can be used to perform image feature recognition, so that the model outputs the corresponding agarwood processing quality level. Therefore, the model can be combined to perform image recognition and fingerprint identification on massive agarwood slice images, greatly improving the efficiency and accuracy of agarwood identification.

[0073] The implementation principle of the present invention will be described in detail below.

[0074] As an optional implementation scheme of the present application, optionally, S1, collecting an image of agarwood song slice, comprises:

[0075] Batch collect the agarwood slice images of a number of agarwood slices and pre-process them;

[0076] Slicing the agarwood song slice image to obtain image slices corresponding to a number of agarwood song slices;

[0077] Check the position correspondence between the image slice and the corresponding agarwood song slice, and record the position information of the image slice.

[0078] The image acquisition module for collecting images of agarwood song slices can use industrial cameras to take pictures, etc., and can batch collect image data of several agarwood song slices after slicing (collect once, one image contains images of several slices) and upload them to the background server.

[0079] The corresponding image is sliced ​​in the background to generate a single slice image of the corresponding agarwood song slice. Then, each slice is traversed and imported into the model for detection.

[0080] The specific implementation steps of slicing an image to obtain a number of image slices can refer to the following steps:

[0081] 1. Prepare the image

[0082] Load the image: Use image processing software (such as Python's PIL / Pillow library, OpenCV library, or professional image processing software such as Photoshop and GIMP) to load the original image to be sliced.

[0083] Confirm the image format: Make sure the image format is suitable for slicing. Common formats include JPEG, PNG, TIFF, etc.

[0084] 2. Determine the slice parameters

[0085] Slice size: Determine the size of each slice (such as width and height) as required. It can be a fixed size or a ratio relative to the original image.

[0086] Number of slices: Calculates or determines the number of slices that need to be generated, which is usually based on the slice size and the dimensions of the original image.

[0087] Overlap (optional): If necessary, you can set the overlap area between slices to ensure smooth stitching or fusion of slices in subsequent processing.

[0088] 3. Perform slicing operations

[0089] Write code or use tools:

[0090] In a programming environment such as Python, you can use loops and functions from image processing libraries to iterate over the image and generate slices.

[0091] In image processing software, batch processing or scripting functions can be used to automate the sectioning process.

[0092] Slice Save: Save each slice as a separate file, usually in a specified folder for later processing.

[0093] 4. Verify the slicing results

[0094] Check the slices: Review the generated slices one by one to make sure they correctly reflect the content of the original image and that there are no missing or duplicated parts.

[0095] Adjust parameters (if necessary): If you find that the slicing result does not meet your expectations, you can return to step 2 to adjust the slicing parameters and re-execute the slicing operation.

[0096] 5. Subsequent processing

[0097] Slice processing: further process the generated slices, such as compression, format conversion, adding watermarks, etc.

[0098] Slice stitching: If necessary, slices can be stitched back into the original image or a new image layout can be created in subsequent steps. That is, after the model outputs the quality level, the quality level is marked on the corresponding slice image, and then the slices are stitched into one image in the original order, so that the quality level of each slice can be displayed in batches on one image for easy management and statistics.

[0099] For the subsequent identification and marking of the unit price, the marking and splicing can also be performed in the above manner to obtain an image marked with the unit price of each slice, which is convenient for counting the unit price of each slice.

[0100] like Figure 2 As shown, as an optional implementation scheme of the present application, optionally, the method for generating the quality inspection recognition model includes:

[0101] Collect and pre-process several agarwood slice images;

[0102] Traversing each of the agarwood slice images, and using CNN to extract texture features in the agarwood slice images, including glossiness, texture spatial distribution, and texture morphology;

[0103] Performing feature annotation on the texture features, wherein the annotation information includes a quality grade determined based on the glossiness, texture spatial distribution, and texture morphology;

[0104] Constructing a feature set based on the labeled texture features;

[0105] Dividing the feature set into a training set and a validation set according to a preset ratio;

[0106] Input the training set into a preset CNN model to perform feature training and learning to construct the quality inspection recognition model;

[0107] The recognition performance of the quality inspection recognition model is verified using the verification set:

[0108] If the verification is passed, the quality inspection and recognition model is deployed on the server and used for texture feature quality recognition of the agarwood song slice image;

[0109] If the verification fails, repeat the above steps to rebuild the quality inspection recognition model.

[0110] Convolutional neural networks can better identify image features and extract image features from images. For the network structure of convolutional neural networks and their model training and application principles, please refer to the technical description of CNN in the prior art.

[0111] The present invention can collect and pre-process several agarwood song slice images in the early stage. The pre-processing can be operations such as cleaning and screening. Images with uneven gloss or texture spatial distribution, low gloss, etc. can be deleted. Subsequently, a convolutional neural network, that is, a CNN network, can be used to extract texture features in the agarwood song slice image. Among the extracted texture features, the color, gloss, texture spatial distribution, and texture morphology in the current agarwood song slice image can be extracted. The texture size and spacing between textures in the slice image can also be calculated and annotated through annotation tools. Features, etc. After extracting the features, the features are annotated. The administrator can comprehensively evaluate the quality level of the current slice based on the extracted texture features, and annotate the corresponding quality level according to industry standards.

[0112] When marking the quality level, the quality level of the corresponding features can be marked according to industry standards and quality inspection conditions. After marking, a feature set consisting of several texture features can be generated, which can be divided into a training set and a verification set according to a preset ratio (such as 8:2). The CNN model is used to perform feature training and learning on several texture features in the training set to construct a quality inspection recognition model. After verification, the model can be deployed in the background to perform quality recognition on the agarwood slice images in subsequent processing, perform corresponding quality level recognition and predict output.

[0113] The general construction steps of the CNN (Convolutional Neural Network) model include the following key parts:

[0114] Data preparation and preprocessing:

[0115] Data Collection: Collect sufficient training data and validation / testing data from reliable sources.

[0116] Data preprocessing: Perform preprocessing operations such as normalization, cropping, scaling, rotation, and flipping on image data to unify the data format and enhance the generalization ability of the model.

[0117] Data augmentation (optional): Increase the diversity of training data through random transformations, such as random cropping, random brightness adjustment, etc.

[0118] Data partitioning: Divide the dataset into training set, validation set, and test set (if the data size is large enough).

[0119] Model definition:

[0120] Choose a framework: Choose a suitable deep learning framework, such as TensorFlow, PyTorch, etc.

[0121] Build the network structure: define the hierarchical structure of CNN, including input layer, convolution layer, activation layer (such as ReLU), pooling layer (such as maximum pooling), fully connected layer, output layer, etc.

[0122] Set parameters: Set parameters for each layer, such as convolution kernel size, step size, padding method, activation function type, pooling window size, number of neurons in the fully connected layer, etc.

[0123] Loss function and optimizer:

[0124] Select loss function: Select a suitable loss function according to the task type (such as classification, regression), such as cross entropy loss, mean square error, etc.

[0125] Select optimizer: Select an optimization algorithm, such as SGD, Adam, RMSprop, etc., and set hyperparameters such as learning rate and momentum.

[0126] Model training:

[0127] Data loading: Use a data loader (such as DataLoader) to load the training data into memory in batches.

[0128] Forward propagation: Calculate the model's prediction value on the current batch of data.

[0129] Calculate loss: Use the loss function to calculate the difference between the predicted value and the true value.

[0130] Backpropagation: Calculate the gradient based on the loss value and update the model parameters through the optimizer.

[0131] Iterative training: Repeat the above steps until the predetermined training rounds are reached or other stopping conditions are met.

[0132] Monitor the training process: record the loss value, accuracy and other indicators during the training process, and use the validation set to evaluate the model performance.

[0133] Model Validation and Evaluation:

[0134] Verify the model: Evaluate the performance of the model on the validation set to check whether the model is overfitting or underfitting. For example, the F1 score is widely used to evaluate the performance of segmentation models, especially for tasks such as tumor detection. The F1 score can comprehensively consider the accuracy and recall rate to help doctors understand the performance of the model more comprehensively. The administrator can complete the verification.

[0135] Adjust hyperparameters: Adjust hyperparameters such as model structure, learning rate, batch size, etc. according to the verification results, and retrain the model.

[0136] Test model (optional): Evaluate the final performance of the model on an independent test set to obtain unbiased model evaluation results.

[0137] Model deployment and optimization (optional):

[0138] Model export: Export the trained model to a deployable format, such as TensorFlow SavedModel, PyTorch model file, etc.

[0139] Model optimization: Perform optimization operations such as quantization and pruning on the model to reduce the model size and inference time.

[0140] Deploy the model: Deploy the model to the target platform (such as server, mobile device, embedded device, etc.) and put it into practical use.

[0141] The above are the general steps for building a CNN model. The specific implementation may vary depending on task requirements, data characteristics, and hardware conditions.

[0142] As an optional implementation scheme of the present application, optionally, S2, inputting the agarwood song slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood song slice image and outputs the corresponding quality level, including:

[0143] Inputting the agarwood song slice image into the quality inspection and recognition model;

[0144] Identify the texture features in the agarwood song slice image through the quality inspection recognition model, and predict and output the quality level corresponding to the texture features;

[0145] The quality grade is bound to the corresponding agarwood song slice image and written into the corresponding image table to obtain an agarwood song quality identification table.

[0146] In the step of identifying the processing quality of agarwood, the slice images of agarwood can be collected and uploaded to the background. After the slices are preprocessed by the background, each slice is input into the quality inspection recognition model in turn, and the model recognizes the image texture features of each agarwood slice, and the model predicts and outputs the corresponding quality grade.

[0147] Subsequently, the quality level output by the model can be bound to each slice, or the quality level can be marked on the corresponding agarwood slice image, and each slice can be written into the corresponding image table, so as to batch generate the quality identification table of agarwood, which can be traversed and browsed to obtain the quality identification results of each slice, which is convenient for quality statistics. The serial number of the corresponding image slice and the quality level marking can be marked and bound according to its position or preset arrangement information.

[0148] If the quality inspection level of the corresponding slice is found to be unqualified during the quality inspection process, an alarm signal for the corresponding agarwood slice can be generated by the background and fed back to the quality inspection terminal. The quality inspection terminal can be a PDA terminal, etc., to remind the administrator to remove the unqualified agarwood slice. Therefore, quality supervision can be carried out in a timely manner and unqualified agarwood slices can be removed.

[0149] The removal of unqualified agarwood slices can be replaced by a robot. Specifically, the background can generate the movement trajectory of the robot according to the location information of the identified unqualified slices, control the robot to go to the location of the unqualified slices, grab and remove them, and the corresponding robot application control software (install the application software according to the constructed device) can be deployed in the background. The generation and control of the movement trajectory of the robot can refer to the movement principle of the existing robot, and this embodiment is not limited.

[0150] As an optional implementation scheme of the present application, optionally, the annotation information further includes:

[0151] The unit price is determined based on the glossiness, texture spatial distribution, and texture morphology.

[0152] As an optional implementation scheme of the present application, optionally, S2, inputting the agarwood song slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood song slice image and outputs the corresponding quality level, including:

[0153] Inputting the agarwood song slice image into the quality inspection and recognition model;

[0154] Identify the texture features in the agarwood slice image through the quality inspection recognition model, and predict and output the unit price corresponding to the texture features;

[0155] The unit price is bound to the corresponding agarwood song slice image and written into the corresponding image table to obtain an agarwood song unit price prediction table.

[0156] In addition to the above-mentioned quality level prediction output, the present invention can also allow the administrator to comprehensively evaluate and mark the value of the corresponding slice, that is, the unit price, based on the texture features of the identified slice image when marking the features. The unit prices of various slices can be marked separately according to the differences in their texture features. In this way, the model can predict and output the unit price of the corresponding slice at the same time after identifying the texture features of the corresponding slice.

[0157] According to the above-mentioned quality grade recording method, the value of each slice, that is, the unit price, can be predicted in batches. Then, the unit price of each slice is bound to each slice image and written into the table, and the unit price prediction table of each agarwood slice in this batch can be obtained, which is convenient for subsequent administrators to count and evaluate the value of the agarwood slices produced and processed. Therefore, it is convenient for enterprise users to estimate the value of slices at the same time as quality appraisal.

[0158] Therefore, the present invention uses big data technology to perform feature training and learning on the texture features of agarwood song slice images, construct a quality inspection and recognition model and apply it to identify the quality level of agarwood song slice images. The AI ​​model can be used to identify the texture features of agarwood song slice images and output the corresponding quality level, and the AI ​​technology can be used to perform image feature recognition, so that the model outputs the corresponding agarwood song processing quality level. Therefore, the model can be combined to perform image recognition and fingerprint identification on massive agarwood song slice images, greatly improving the efficiency and accuracy of agarwood song identification. The adoption of this solution can lower the entry threshold and requirements for identification, realize large-scale quality inspection, and improve the processing quality outbound standards of agarwood song.

[0159] It should be noted that although CNN is used as an example, those skilled in the art will understand that the present disclosure should not be limited thereto. In fact, users can flexibly set the training model according to the actual application scenario, as long as the technical functions of the present application can be realized according to the above technology.

[0160] Obviously, those skilled in the art should understand that the implementation of all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. Those skilled in the art can understand that the implementation of all or part of the processes in the above-mentioned embodiments can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated as: HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above-mentioned types of memory.

[0161] Example 2

[0162] like Figure 3 As shown, based on the implementation principle of Example 1, on the other hand, the present application proposes a quality control system for agarwood production based on image recognition technology, including:

[0163] An image acquisition module is used to collect images of agarwood slices and input them into the agarwood quality inspection module;

[0164] Agarwood quality inspection module, used for inputting the agarwood slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood slice image and outputs the corresponding quality grade;

[0165] The quality control module is used to determine whether the quality level of the current agarwood slices meets the standards:

[0166] If the standard is met, the next agarwood slice will be quality inspected;

[0167] Otherwise, an alarm signal is issued;

[0168] An alarm module, used for responding to the alarm signal;

[0169] The image acquisition module, agarwood song quality inspection module, quality control module and alarm module are communicatively connected in sequence.

[0170] The module functions and interactions of the above system can be understood in conjunction with the corresponding steps of Example 1, and will not be described in detail in this example.

[0171] The modules or steps of the present invention described above can be implemented by a general-purpose computing system, they can be concentrated on a single computing system, or distributed on a network composed of multiple computing systems, and optionally, they can be implemented by a program code executable by a computing system, so that they can be stored in a storage system and executed by the computing system, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0172] Example 3

[0173] like Figure 4 As shown, further, in another aspect, the present application also proposes an electronic device, including:

[0174] processor;

[0175] a memory for storing processor-executable instructions;

[0176] Among them, the processor is configured to implement the agarwood song production quality control method based on image recognition technology described in Example 1 when executing the executable instructions.

[0177] The electronic device of the embodiment of the present disclosure includes a processor and a memory for storing processor executable instructions. The processor is configured to implement the agarwood song production quality control method based on image recognition technology described in the above embodiment 1 when executing the executable instructions.

[0178] Here, it should be noted that the number of processors can be one or more. At the same time, the electronic device of the embodiment of the present disclosure may also include an input system and an output system. Among them, the processor, memory, input system and output system may be connected through a bus or in other ways, which are not specifically limited here.

[0179] The memory, as a computer-readable storage medium, can be used to store software programs, computer executable programs and various modules, such as the program or module corresponding to the agarwood song production quality control method based on image recognition technology in the embodiment of the present disclosure. The processor executes various functional applications and data processing of the electronic device by running the software programs or modules stored in the memory.

[0180] The input system can be used to receive input numbers or signals. The signal can be a key signal related to user settings and function control of the device / terminal / server. The output system can include display devices such as display screens.

[0181] The embodiments of the present disclosure have been described above, and the above description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and changes will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The selection of terms used herein is intended to best explain the principles of the embodiments, practical applications, or technical improvements in the market, or to enable other persons of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method for controlling the quality of agarwood production based on image recognition technology, characterized in that: The steps include: S1, collecting images of agarwood slices; S2, inputting the agarwood song slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood song slice image and outputs the corresponding quality grade; S3, judging whether the quality level of the current agarwood slices meets the standard: If the standard is met, the next agarwood slice will be quality inspected; Otherwise, an alarm signal is issued.

2. According to claim 1, the method for controlling the quality of agarwood production based on image recognition technology is characterized in that: S1. Collecting slice images of agarwood song, including: Batch collect and pre-process the agarwood slice images of a number of agarwood slices; Slicing the agarwood song slice image to obtain image slices corresponding to a number of agarwood song slices; Check the position correspondence between the image slice and the corresponding agarwood song slice, and record the position information of the image slice.

3. The method for controlling the quality of agarwood production based on image recognition technology according to claim 1, characterized in that: The method for generating the quality inspection recognition model comprises: Collect and pre-process several agarwood slice images; Traversing each of the agarwood slice images, and using CNN to extract texture features in the agarwood slice images, including glossiness, texture spatial distribution, and texture morphology; Performing feature annotation on the texture features, wherein the annotation information includes a quality grade determined based on the glossiness, texture spatial distribution, and texture morphology; Constructing a feature set based on the labeled texture features; Dividing the feature set into a training set and a validation set according to a preset ratio; Input the training set into a preset CNN model to perform feature training and learning to construct the quality inspection recognition model; The recognition performance of the quality inspection recognition model is verified using the verification set: If the verification is passed, the quality inspection and recognition model is deployed on the server and used for texture feature quality recognition of the agarwood song slice image; If the verification fails, repeat the above steps to rebuild the quality inspection recognition model.

4. The method for controlling the quality of agarwood production based on image recognition technology according to claim 3 is characterized in that: The annotation information also includes: The unit price is determined based on the glossiness, texture spatial distribution, and texture morphology.

5. According to claim 3, the method for controlling the quality of agarwood production based on image recognition technology is characterized in that: S2, inputting the agarwood song slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood song slice image and outputs the corresponding quality level, including: Inputting the agarwood song slice image into the quality inspection and recognition model; Identify the texture features in the agarwood slice image through the quality inspection recognition model, and predict and output the quality level corresponding to the texture features; The quality grade is bound to the corresponding agarwood song slice image and written into the corresponding image table to obtain an agarwood song quality identification table.

6. The method for controlling the quality of agarwood production based on image recognition technology according to claim 4 is characterized in that: S2, inputting the agarwood song slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood song slice image and outputs the corresponding quality level, including: Inputting the agarwood song slice image into the quality inspection and recognition model; Identify the texture features in the agarwood slice image through the quality inspection recognition model, and predict and output the unit price corresponding to the texture features; The unit price is bound to the corresponding agarwood song slice image and written into the corresponding image table to obtain an agarwood song unit price prediction table.

7. A quality control system for agarwood production based on image recognition technology, used to implement the quality control method for agarwood production based on image recognition technology as described in any one of claims 1 to 6, characterized in that: include: An image acquisition module is used to collect images of agarwood slices and input them into the agarwood quality inspection module; Agarwood quality inspection module, used for inputting the agarwood slice image into a preset quality inspection recognition model, and the quality inspection recognition model recognizes the texture features of the agarwood slice image and outputs the corresponding quality grade; The quality control module is used to determine whether the quality level of the current agarwood slices meets the standards: If the standard is met, the next agarwood slice will be quality inspected; Otherwise, an alarm signal is issued; An alarm module, used for responding to the alarm signal; The image acquisition module, agarwood song quality inspection module, quality control module and alarm module are communicatively connected in sequence.

8. An electronic device, characterized in that include: processor; a memory for storing processor-executable instructions; Wherein, the processor is configured to implement the agarwood song production quality control method based on image recognition technology as described in any one of claims 1-6 when executing the executable instructions.

Images