Method and system for visual recognition and state characterization of the cap hole of morchella during drying process

Abstract

This invention discloses a method and system for visual recognition and state characterization of cap pores during the drying process of morel mushrooms. The method first constructs a region of interest (ROI) on the morel cap, then performs instance-level recognition and segmentation of individual pores on the cap surface to obtain accurate pore segmentation results. Next, based on the segmentation results, it extracts pore structural features and cap morphological and color features from the ROI. Subsequently, it fuses these three types of features to construct a multi-dimensional feature vector, establishing a unified analytical framework for local pores and the overall cap state. Finally, it inputs the multi-dimensional feature vector into a pre-trained regression model to output the morel mushroom drying state characterization results. This method overcomes the limitations of traditional techniques in treating pores holistically, effectively avoiding problems such as pore adhesion and missed detection, improving the robustness of feature characterization and the consistency of quantitative results, achieving accurate monitoring of the morel mushroom drying process, and providing reliable technical support for drying state characterization and quality control.

Description

Technical Field

[0001] This invention belongs to the field of agricultural product processing monitoring and image processing technology, specifically relating to a method and system for visual recognition and state characterization of cap pores in morel mushroom drying process monitoring. Background Technology

[0002] Morel mushrooms, as an edible fungus of significant economic value, exhibit a typical porous and irregular structure on their cap surface, characterized by numerous pores of varying sizes and complex spatial distribution. During the drying process of morel mushrooms, as moisture migrates and is removed, the cap tissue undergoes continuous structural rearrangement, including overall shrinkage, local collapse, deformation of pore boundaries, and changes in the connectivity and aggregation between pores. These changes in apparent structure reflect, to some extent, the stage characteristics and tissue state of the drying process.

[0003] However, existing methods for monitoring and evaluating the drying process of morel mushrooms mostly focus on controlling drying time, measuring water loss rate, or making empirical judgments about overall appearance. These methods struggle to accurately depict the dynamic evolution of the complex porous structure on the cap surface and are poorly adaptable to individual sample differences. In image analysis techniques, some methods have attempted to analyze the morel mushroom surface using image segmentation, grayscale statistics, or texture features. However, these methods typically treat the porous structure as an overall texture or regional feature, lacking the ability to identify and model individual pores as independent structural units. When pore sizes vary greatly, shapes are irregular, or distributions are dense, problems such as pore adhesion, missed detections, or misjudgments can easily occur, leading to insufficient stability and consistency in quantitative results. Furthermore, some processing methods based on empirical thresholds or simple rules are sensitive to imaging conditions and sample states, limiting their versatility and robustness, and failing to meet the needs of continuous monitoring of the drying process and engineering applications. Therefore, existing technologies still have significant shortcomings in the identification and quantitative analysis of the cap porous structure during the morel mushroom drying process. Summary of the Invention

[0004] The purpose of this invention is to solve the problems of difficulty in identifying fungal cap pores at the instance level and unreliable quantitative analysis and monitoring in the prior art, and to provide a method and system for visual identification and state characterization of fungal cap pores during the drying process of morel mushrooms.

[0005] To achieve the above objectives, the present invention employs the following technical solution: The present invention proposes a method for visual identification and state characterization of cap pores during the morel drying process, comprising the following steps: Construct the region of interest (ROI) of the morel mushroom cap, and based on the ROI, perform instance-level identification and segmentation of individual pores on the cap surface to obtain pore segmentation results; Based on the hole segmentation results, hole structure features are extracted, and mushroom cap morphology and color features are extracted based on the region of interest of the mushroom cap. The features of pore structure, cap morphology and color are fused to construct a multi-dimensional feature vector; By inputting multidimensional feature vectors into a pre-trained morel drying state prediction regression model, the model outputs the morel drying state characterization results, thereby enabling the monitoring of the morel drying process.

[0006] The present invention proposes a system for visual recognition and state characterization of cap pores for monitoring the drying process of morel mushrooms, comprising: A hole segmentation module is used to construct the region of interest of morel mushroom caps, and based on the region of interest of morel mushroom caps, to perform instance-level identification and segmentation of individual holes on the surface of the caps to obtain hole segmentation results. The feature extraction module is used to extract hole structure features based on hole segmentation results, and extract mushroom cap morphology features and mushroom cap color features based on the region of interest of the mushroom cap. The feature fusion module is used to fuse the features of the pore structure, the morphology of the mushroom cap, and the color of the mushroom cap to construct a multi-dimensional feature vector. The result prediction module is used to input multidimensional feature vectors into a pre-trained morel drying state prediction regression model and output the morel drying state characterization results to realize the monitoring of the morel drying process.

[0007] A terminal device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for visual recognition and state characterization of cap pores for monitoring the drying process of morel mushrooms.

[0008] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of a method for visually identifying and characterizing the state of cap pores during the monitoring of the morel drying process.

[0009] Compared with the prior art, the present invention has the following beneficial effects: To address the problems existing in the current technology, a method for visual recognition and state characterization of cap pores in morel mushroom drying process monitoring is proposed. This method can identify cap pores at the instance level in the drying process monitoring scenario for complex porous surfaces, and perform stable quantitative analysis of structural features such as the number, area, morphology and spatial distribution of pores. This provides a more reliable technical means for state characterization and quality control of morel mushroom drying process. Specifically, firstly, a region of interest (ROI) is constructed for the morel mushroom cap. Instance-level identification and segmentation of individual pores on the cap surface are then performed, overcoming the limitations of existing technologies that treat pores as overall texture or regional features. This effectively avoids problems such as adhesion, missed detection, and misjudgment caused by large differences in pore scale and dense distribution. Subsequently, pore structural features are extracted based on the pore segmentation results. Simultaneously, cap morphological and color features are extracted from the ROI. Through multi-dimensional features, the correlation information of pore evolution, cap shrinkage and collapse, and color changes during the drying process is comprehensively captured, overcoming the deficiency of single features in depicting complex drying states. Then, multi-dimensional feature vectors are constructed through feature fusion, establishing a unified analytical framework for local pores and the overall cap state. This improves the completeness and robustness of feature representation and reduces sensitivity to imaging conditions and sample state. Finally, the multi-dimensional feature vectors are input into a pre-trained regression model, outputting accurate morel mushroom drying state characterization results. This replaces the inefficient traditional methods that rely on drying time, water loss rate measurement, or empirical judgment, achieving accurate monitoring of the drying process and providing reliable technical support for drying process state characterization and quality control.

[0010] Preferably, by transforming the overall texture features of the cap pores into instance-level structural units, and combining this with a comprehensive analysis of the overall morphology and color features of the cap, a multidimensional quantitative characterization of the changes in the apparent structure of morel mushrooms during the drying process is achieved. This method can stably extract key features under complex conditions of dense cap pores, irregular morphology, and continuous color changes, improving the accuracy and interpretability of the drying process monitoring results. Furthermore, the method of this invention does not rely on dedicated imaging equipment, has a clear process, strong adaptability, and good engineering implementation and application value. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of the method for visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms according to the present invention.

[0013] Figure 2This is a detailed flowchart of the method for visual recognition and state characterization of fungal cap pores according to the present invention.

[0014] Figure 3 This is a schematic diagram of the light-shielding imaging device used for acquiring images of the surface of morel mushroom caps in an embodiment of the present invention.

[0015] Figure 4 This is a schematic diagram of RGB images of the surface of morel caps at different drying stages in an embodiment of the present invention.

[0016] Figure 5 This is a schematic diagram illustrating the process of image preprocessing and region of interest (ROI) construction for morel caps in an embodiment of the present invention.

[0017] Figure 6 This is a schematic diagram of the instance-level identification and segmentation results of pores on the surface of morel caps in an embodiment of the present invention.

[0018] Figure 7 This is a schematic diagram illustrating the extraction and visualization of the pore structure features of morel mushroom caps at different drying stages in an embodiment of the present invention.

[0019] Figure 8 This is a schematic diagram illustrating the distribution and evolution of the color characteristics of the morel cap surface at different drying stages in an embodiment of the present invention.

[0020] Figure 9 This is a schematic diagram comparing the predicted drying state of morel mushrooms based on a regression model with the actual values ​​in an embodiment of the present invention.

[0021] Figure 10 This is a diagram of the system for visual identification and state characterization of cap pores for monitoring the morel drying process according to the present invention.

[0022] Figure 11 This is a schematic diagram of the structure of an electronic device according to the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0024] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0025] The present invention will now be described in further detail with reference to the accompanying drawings: Example 1 To address the shortcomings of existing technologies that rely primarily on drying time, quality changes, or empirical judgments of overall appearance for monitoring the drying process of morel mushrooms, which struggles to provide precise, stable, and quantifiable characterization of the complex structural changes on the cap surface, particularly the difficulty in establishing a unified analytical framework across multiple dimensions such as pore structure, overall morphological changes, and color evolution, leading to insufficient accuracy and consistency in drying status monitoring, this invention proposes a method for visual recognition and status characterization of cap pores during morel mushroom drying. This invention transforms the overall texture features of cap pores into independently identifiable instances, achieving instance-level recognition and segmentation of pore structures within the cap area. Based on this, combined with a comprehensive analysis of overall cap morphological and color parameters, it achieves multidimensional quantitative characterization of the cap's surface structure evolution during the drying process, thus providing a reliable technical means for drying process monitoring and status assessment. The proposed method for visual recognition and status characterization of cap pores during morel mushroom drying is described below. Figure 1 As shown, it includes the following steps: S1. Construct the region of interest (ROI) of the morel mushroom cap. Based on the ROI, perform instance-level identification and segmentation of individual holes on the cap surface to obtain hole segmentation results. The construction of the region of interest for the morel mushroom cap specifically involves: The original cap surface image is converted to grayscale to obtain a grayscale image of the cap surface. The grayscale image is then inverted and subjected to strong Gaussian smoothing. After smoothing, an adaptive threshold segmentation method is used for binarization to obtain an initial binary image of the cap region. Morphological closing operation is performed on the binary image to obtain a mask image of the cap region, thus constructing the region of interest of the morel cap.

[0026] The process involves instance-level identification and segmentation of individual pores on the surface of the morel cap, based on the region of interest of the cap, to obtain the pore segmentation results. Specifically: The region of interest of the morel mushroom cap is divided into several overlapping or non-overlapping local sub-regions according to a preset size. The hole structure is independently identified in each sub-region. During the identification process, each identified hole instance is independently marked and a corresponding instance segmentation mask is generated to obtain the hole instance-level segmentation result.

[0027] S2. Extract hole structure features based on hole segmentation results, and extract mushroom cap morphology and color features based on the region of interest of the mushroom cap. Based on the hole segmentation results, hole structure features are extracted, specifically: The pore structure features include pore area, the ratio of pore area to the projected area of ​​the cap, number of pores, distribution of equivalent diameter of pores, spatial distribution density between pores, and pore connectivity features. By comparing and analyzing the pore structure features at different drying stages, the evolution trend of pores from the initial state to the shrinkage, deformation and rearrangement state during the drying process is quantitatively described.

[0028] Based on the region of interest of the cap, the morphological features of the cap are extracted, specifically: The outline boundary of the mushroom cap is extracted, and the projected area and perimeter of the mushroom cap are calculated. The principal axis and secondary axis direction of the mushroom cap outline are analyzed, and the principal axis length and secondary axis length of the mushroom cap are calculated respectively. By the change relationship between the principal axis and secondary axis length, the anisotropic shrinkage behavior of the mushroom cap during the drying process is characterized.

[0029] Based on the region of interest of the mushroom cap, the color features of the mushroom cap are extracted, specifically: Based on the region of interest (ROI) of the mushroom cap, statistical analysis was performed on the pixel color information on the cap surface, and the color space was converted from RGB to CIE. Color space; specifically, within the region of interest of the mushroom cap, brightness parameters are extracted pixel by pixel. and chromaticity parameters Furthermore, the color parameters under the same drying stage were statistically summarized to obtain the color distribution characteristics of the cap surface at that stage.

[0030] S3. The features of pore structure, cap morphology and color are fused to construct a multi-dimensional feature vector; S4. Input the multidimensional feature vector into the pre-trained morel drying state prediction regression model, output the morel drying state characterization result, and realize the monitoring of the morel drying process.

[0031] The pre-trained regression model for predicting the drying state of morel mushrooms is as follows: Morel mushroom samples at different drying stages were used as training samples. The mapping relationship between the comprehensive feature vector and the corresponding drying state was learned through regression modeling. After the model training was completed, the trained regression model was applied to the test samples to predict the drying state of morel mushrooms and obtain the corresponding prediction results. The prediction results were compared and analyzed with the actual drying state and displayed in a visualization manner to evaluate the prediction performance of the model.

[0032] The method is described in detail below: S1. Obtain images of the cap surface of morel mushroom samples at different drying stages, preprocess the cap surface images, extract the cap region, and construct the cap region of interest for subsequent analysis. Image preprocessing includes at least one of image scale unification, brightness correction, background suppression, noise removal, and region connectivity correction to improve the stability and consistency of the extraction process of fungal cap pore, morphology, and color parameters.

[0033] S2. Based on the region of interest of the cap, identify the holes on the surface of the cap, segment each hole as an independent instance, and obtain the instance-level segmentation result of the cap holes. The mushroom cap hole identification adopts an instance-level processing method, which distinguishes adjacent or similar hole structures as independent instances, avoiding treating holes as continuous texture areas for overall processing, thereby improving the completeness of hole identification and the reliability of quantitative results.

[0034] S3. Based on the instance-level segmentation results, extract one or more pore structure parameters from the pore quantity, pore area, pore size distribution, pore boundary features, pore spatial distribution, and pore connectivity of the cap pores to form a quantitative feature set of cap pores; S4. Based on the region of interest of the cap, analyze the overall morphology of the cap and extract morphological parameters that reflect the geometric changes of the cap. The morphological parameters include one or more of the following: cap projected area, perimeter, length, width and their variation relationship. S5. Based on the region of interest of the mushroom cap, analyze the color information of the mushroom cap surface and extract color parameters that reflect the color evolution characteristics during the drying process. The color parameters include one or more of the following: brightness parameters, chromaticity parameters, and color distribution characteristics. The pore structure parameters, morphological parameters, and color parameters can be used individually or in combination to characterize the evolution of the cap surface structure during the drying process of morel mushrooms.

[0035] S6. The pore structure parameters, morphological parameters and color parameters obtained at different drying stages are comprehensively processed to construct a quantitative characterization result of the changes in the appearance characteristics of the cap with the drying process, which is used to characterize the state changes of morel mushrooms during the drying process.

[0036] The present invention also provides an application of the method in monitoring the drying process of morel mushrooms. By comprehensively analyzing the pore structure, overall morphology and color changes of the cap, the state of the drying process can be monitored and evaluated.

[0037] The following is combined Figures 2 to 9 This invention provides a method for identifying and quantitatively analyzing cap pores for monitoring the drying process of morel mushrooms, and its application process.

[0038] like Figure 2This is a flowchart illustrating the method for identifying and quantitatively analyzing cap pores in morel mushrooms for monitoring the drying process provided by this invention. The method processes images of the morel mushroom cap surface to achieve a comprehensive analysis of pore structure, overall morphology, and color characteristics, characterizing the state changes during the drying process. The process includes, but is not limited to, the following steps: Step 1: Obtain raw images of the cap surface of morel mushroom samples at different drying stages, and preprocess and suppress the background of the raw images to obtain standardized image data for analysis. Step 2: Based on the preprocessed image, the cap region is automatically identified and extracted to construct the region of interest (ROI) of the cap, so as to eliminate the interference of non-target background regions on subsequent analysis; Step 3: Enhance the pore boundaries and highlight local structures in the region of interest of the cap to improve the distinguishability of the pore outline and internal structural features. Step 4: Based on the enhanced cap image, perform instance-level identification and segmentation of the pores on the cap surface, label each pore as an independent structural instance, and obtain the instance-level segmentation result of the cap pores. Step 5: Based on the instance-level segmentation results, extract pore structure parameters such as pore area ratio, pore connectivity, pore boundary complexity, and pore aggregation degree from the cap pores to form a set of pore structure features. Step 6: Based on the region of interest of the cap, analyze the overall morphology of the cap and extract morphological feature parameters such as the projected area, perimeter, principal axis length and secondary axis length of the cap; Step 7: Based on the region of interest of the mushroom cap, analyze the color information of the mushroom cap surface and extract the brightness parameter. and chromaticity parameters Color feature parameters; Step 8: Perform multi-feature fusion processing on the pore structure feature parameters, morphological feature parameters and color feature parameters to construct a comprehensive feature vector for characterizing the drying state of morel mushrooms; Step 9: Establish a drying state regression model based on the comprehensive feature vector, and use the regression model to predict the drying state of morel mushrooms to obtain the corresponding quantitative characterization results of the drying process.

[0039] The analysis focuses on the surface structure of morel mushroom caps, including but not limited to the morphology, distribution, overall shape, and color changes of the cap's pores. This information reflects the structural evolution and state changes of morels during the drying process. Because the surface of morel mushroom caps exhibits a typical porous and irregular structure with densely packed pores of significant scale variations, and with surface wrinkles and pore boundaries highly interwoven in space, it is easily affected by background interference, uneven lighting, and shadow occlusion under conventional imaging conditions. This leads to blurred pore boundaries and loss of local structural details, thus affecting the accuracy and stability of subsequent pore identification and quantitative analysis. Therefore, establishing a stable and repeatable image acquisition environment is fundamental for the reliable identification of morel mushroom cap pore structures.

[0040] To address the aforementioned problems, this invention employs a combination of fixed artificial lighting and a light-blocking imaging structure during the image acquisition stage to construct a relatively enclosed and stable imaging environment. For example... Figure 3 As shown, the left side is a magnified view of sample placement, clearly revealing the cap structure. The right side shows a light-shielded imaging chamber equipped with fixed artificial illumination to provide uniform and stable light. A top-view camera on the top of the chamber is responsible for acquiring images of the morel mushrooms. This invention constructs a light-shielded imaging chamber during image acquisition, with a fixed artificial light source inside to provide uniform illumination to the morel mushroom samples, avoiding the influence of ambient light variations on the imaging results. The artificial light source is preferably positioned above and to the side of the imaging chamber, using multi-angle illumination to reduce the shadow effect inside and at the edges of the cap pores, thereby improving the discernibility of the pore outline and internal structure.

[0041] In the specific implementation process, morel mushroom samples at different drying stages are placed on the imaging stage in a natural manner, with the caps facing upwards and exposed to the imaging area. Subsequently, the imaging device is fixedly set at a preset position above the imaging cavity to perform vertical imaging acquisition of the morel mushroom cap surface, obtaining raw image data containing the cap area. The imaging device can be a conventional visible light camera or an industrial camera, without relying on dedicated imaging hardware, thereby improving the engineering applicability of the method.

[0042] After image acquisition, the obtained raw images undergo background suppression and region cropping to remove non-target areas, retaining only the effective image region containing the morel cap. This region serves as input data for subsequent region of interest (ROI) extraction, hole identification, and feature analysis. This approach provides a reliable data foundation for subsequent instance-level hole identification, morphological parameter extraction, and color feature analysis, while ensuring stable and consistent imaging conditions.

[0043] Figure 4 The images shown are RGB images of morel caps at different drying stages obtained in this embodiment of the invention. (a) to (g) show seven morel samples in different states. Figure 4 Based on machine vision processing of the RGB image of the mushroom cap surface shown, a structural candidate region image of the mushroom cap region is obtained. Here, a structural candidate region image refers to a region in the mushroom cap surface image that differs significantly from its surrounding regions in brightness, texture density, or color distribution; such region is then identified as a structural candidate region and segmented.

[0044] Generally, the porous, wrinkled boundary, and locally collapsed areas on the surface of morel caps typically exhibit reduced brightness, denser texture, or significant color changes compared to adjacent tissues in images. Therefore, the candidate structural region images obtained through the above processing steps can include images of the cap porous region and images of local boundary regions related to the porous structure. This step provides a candidate region basis for subsequent porous boundary enhancement, instance-level porous identification, and quantitative analysis.

[0045] In this embodiment of the invention, in order to address the problem of dense pore structure, complex texture and obvious background interference on the surface of morel caps, before performing instance-level pore recognition, the collected RGB images of the cap surface are first preprocessed and the region of interest of the cap is constructed to improve the stability and accuracy of subsequent pore recognition and quantitative analysis.

[0046] Specifically, such as Figure 5 As shown, (a) is the Original ROI, which is the original RGB image of the mushroom cap's region of interest, presenting the texture of the cap and the substrate background; (b) is the Grayscale Image, which converts the original image to grayscale, simplifying color information and focusing on brightness differences; (c) is the Inverted Grayscales, which inverts the grayscale image, making the holes in the mushroom cap brighter areas; (d) is the Strong Gaussian Blur, which applies a strong Gaussian blur to the inverted grayscale image, smoothing noise and weakening the interference of the substrate mesh; (e) is the Otsu Binary Result, which uses the Otsu algorithm to blur the image. Figure 2(f) Morphological closing is performed on the binarized result to fill small holes and connect broken areas, ultimately obtaining a complete and continuous cap region mask. The image preprocessing process includes: grayscale processing of the original RGB image of the cap surface to obtain a grayscale image of the cap surface; further grayscale inversion processing of the grayscale image to enhance the contrast between the hole area and the solid tissue area; then strong Gaussian smoothing processing is applied to the inverted grayscale image to suppress the fine texture and noise interference on the cap surface and retain the overall contour information of the cap. After smoothing, the image is binarized using an adaptive threshold segmentation method, preferably the Otsu threshold segmentation algorithm, to obtain an initial cap region binary image. Since there may be local holes or boundary breaks in the initial binary result, morphological closing processing is further performed on the binary image to fill the internal holes and correct the continuity of the cap boundary, thereby obtaining a cap region mask image with complete structure and clear boundaries. Through the above processing steps, the cap region of morel mushrooms can be automatically identified and extracted without relying on manual annotation, constructing a region of interest (ROI) for the cap and effectively removing interference from the background region on the analysis of hole structure, morphological features, and color features. The obtained cap ROI serves as a unified input basis for subsequent hole boundary enhancement, instance-level hole recognition, and multi-feature extraction.

[0047] Based on the construction of the region of interest (ROI) on the cap and the enhancement of the pore boundaries, instance-level identification and segmentation of pores on the surface of morel caps are performed. For example... Figure 6 As shown, this invention introduces an instance-level recognition strategy based on multi-scale windows within the region of interest (ROI) of the mushroom cap, transforming holes from an overall texture structure into independently identifiable structural instances. Specifically, the ROI of the mushroom cap is divided into several overlapping or non-overlapping local sub-regions according to a preset size, and the hole structure is independently reasoned and recognized within each sub-region to reduce the impact of hole scale differences and local occlusion on the recognition results. This approach improves the completeness of recognizing small-scale holes and holes with complex boundaries while maintaining the continuity of the overall structure. During the instance-level recognition process, each identified hole instance is independently labeled, and a corresponding instance segmentation mask is generated to distinguish the boundary relationships between different holes. Figure 6 As shown, the porous areas on the cap surface are accurately separated as instances, and the adhesion between adjacent pores is effectively suppressed, thus obtaining clear and well-defined instance-level segmentation results for the pores. Through the above instance-level identification and segmentation processing, reliable basic data can be provided for the subsequent quantitative calculation of structural parameters such as pore area, connectivity, spatial distribution, and aggregation characteristics.

[0048] After obtaining instance-level segmentation results of the pores on the surface of morel caps, quantitative calculations of the structural features of the pore instances are performed to characterize the evolution of the pore structure during the drying process. For example... Figure 7 As shown, (a) is the Boundary contour (Perimeter), which outlines the edge of the cap with a red contour, corresponding to the extracted cap perimeter feature, reflecting the overall contour range of the cap and indicating the degree of cap shrinkage during drying; (b) is the Projected cap (area), which fills the cap area with red, corresponding to the extracted cap projected area feature, reflecting the cap's planar coverage range, and is one of the core quantitative indicators of drying deformation; (c) is the Minor axis width, marked by a blue line segment, corresponding to the cap's minor axis width feature; (d) is the Major axis length, marked by a green line segment, corresponding to the cap's major axis length feature. The parameters of these two axes can quantify the morphological proportions of the cap and help determine the cap deformation state caused by drying. This invention uses drying time as a sequence to continuously analyze the pore structure of the same morel sample at different drying stages.

[0049] Specifically, based on an instance-level hole segmentation mask, the corresponding geometric parameters are calculated for each hole instance, and these parameters are further summarized to form the hole structure features at the cap level. These hole structure features include, but are not limited to: hole area, the ratio of hole area to the cap's projected area, the number of holes, the distribution of equivalent hole diameters, the spatial distribution density between holes, and hole connectivity characteristics. By comparing and analyzing these parameters at different drying stages, the evolutionary trend of holes from their initial state to shrinkage, deformation, and rearrangement states during the drying process can be quantitatively described.

[0050] Furthermore, to characterize the spatial heterogeneity of the pore structure on the cap surface, this invention divides the region of interest of the cap into several local sub-regions, and statistically analyzes the pore area ratio, pore density, and pore aggregation degree in each sub-region, thereby obtaining the spatial distribution characteristics of the pore structure. This method not only reflects the overall quantity and scale variation of pores but also characterizes the rearrangement and aggregation behavior of pores in different regions.

[0051] By continuously extracting and visualizing the changes in pore structure features over drying time, this invention achieves a quantitative description of the dynamic evolution of the cap pore structure during the drying process of morel mushrooms, providing a reliable structural feature input for subsequent morphological and color feature fusion analysis and drying state modeling.

[0052] Based on the extraction of the pore structure features of the morel cap, a quantitative analysis of the overall morphological characteristics of the morel cap was performed to characterize the changes in the macroscopic geometric morphology of the cap during the drying process. For example... Figure 8 As shown, using the three-dimensional CIELAB color space, the distribution changes of morel mushroom cap color data points at different drying time points (0 min, 30 min to 360 min) are presented. The color bars at the bottom correspond to the average color at each time point. From the trend, at the initial drying stage (0 min), the color data points are concentrated in... (Brightness) is relatively high. In areas with relatively high temperatures, the mushroom caps are brighter in color; as drying time increases, the data points generally decrease towards L. The changing region migration reflects the process of the mushroom cap's color gradually darkening and its tone continuously turning brown. This invention extracts the overall outline of the mushroom cap based on a pre-constructed region of interest, and then calculates morphological feature parameters reflecting changes in the mushroom cap's geometric shape.

[0053] Specifically, by extracting the outline boundary of the mushroom cap, the projected area and perimeter of the cap are calculated to characterize the overall shrinkage degree and boundary complexity changes of the cap during the drying process. Furthermore, the principal and secondary axes of the cap outline are analyzed, and the lengths of the principal and secondary axes are calculated to reflect the deformation characteristics of the cap in different directions. The relationship between the lengths of the principal and secondary axes can further characterize the anisotropic shrinkage behavior of the cap during the drying process.

[0054] In the above morphological feature extraction process, all morphological parameters are calculated based on the region of interest of the cap at a uniform scale to avoid the influence of imaging scale differences on the results. By comparing and analyzing the morphological feature parameters of the cap at different drying stages, the overall shrinkage, deformation, and contour changes that occur during the evolution of the cap from the initial state to the dry state can be quantitatively described.

[0055] By combining the morphological feature parameters with the aforementioned pore structure features, this invention can comprehensively characterize the changes in the apparent structure of morel mushrooms during the drying process from both macroscopic and microscopic levels, providing important geometric feature inputs for subsequent color feature analysis and drying state modeling.

[0056] Based on the extraction of the structural features of the cap's pores and the overall morphological features, this invention quantitatively analyzes the color features of the morel cap surface to characterize the evolution of cap color over time during the drying process. This invention uses a pre-constructed region of interest (ROI) on the cap to statistically analyze the pixel color information on the cap surface and converts the color space from RGB to a more perceptually consistent one. A color space is used to reduce the impact of differences in imaging conditions on color analysis results. Specifically, within the region of interest of the fungal cap, brightness parameters are extracted pixel by pixel. and chromaticity parameters Furthermore, the color parameters at the same drying stage were statistically summarized to obtain the color distribution characteristics of the cap surface at that stage. This was achieved through three-dimensional... Visualizing the color distribution at different drying times in space can intuitively reflect the overall trend of the mushroom cap color evolving from the initial stage to the darker stage. The parameters gradually decrease with increasing drying time. and The parameters reflect the process of the overall hue of the morel mushroom changing from light brown to dark brown. Furthermore, by calculating the mean and distribution range of the color parameters on the surface of the mushroom cap at each drying time point, a quantitative description of the color features changing with the drying process can be constructed, providing stable color feature input for subsequent multi-feature fusion analysis and drying state modeling. These color features can be used alone to characterize color changes during the drying process, or they can be used in conjunction with pore structure and morphological features to enhance the comprehensive characterization of the morel mushroom's drying state.

[0057] After extracting the structural features of the cap pores, overall morphological features, and color features, a multi-feature fusion process is performed to construct a comprehensive feature vector characterizing the drying state of morel mushrooms. Specifically, the pore structure, morphological, and color feature parameters are uniformly organized, and different features are scaled and standardized. These features are then concatenated according to a preset feature arrangement order to form a comprehensive feature vector containing multi-dimensional information. This multi-feature fusion method organically integrates multiple types of information reflecting changes in the microstructure of the pores, changes in the macroscopic geometric morphology of the cap, and color evolution characteristics, thereby comprehensively describing the apparent structural state of morel mushrooms during the drying process. This multi-feature fusion process provides a unified, stable, and highly representative feature input for the subsequent establishment of a drying state regression model, which is beneficial for improving the accuracy and robustness of drying state prediction.

[0058] Based on the multi-feature fusion processing and the acquisition of a comprehensive feature vector, a regression model for the drying state of morel mushrooms is established to quantitatively predict the drying process. Specifically, morel mushroom samples at different drying stages are used as training samples, and a regression modeling method is used to learn the mapping relationship between the comprehensive feature vector and the corresponding drying state. The regression model can be a linear regression model or a nonlinear regression model, preferably a regression model with strong nonlinear fitting ability to improve the modeling ability of complex feature changes during the drying process.

[0059] After model training is completed, the trained regression model is applied to the test samples to predict the drying state of morel mushrooms and obtain the corresponding prediction results. By comparing and analyzing the prediction results with the actual drying state and displaying them in a visual manner, the model's predictive performance can be evaluated, thereby achieving quantitative characterization and state monitoring of the morel mushroom drying process.

[0060] Through the above description of the embodiments, those skilled in the art can clearly understand that the various embodiments of the present invention can be implemented either by combining software with necessary general-purpose hardware platforms or independently by hardware. Based on this, the parts of the technical solutions of the present invention that make substantial contributions to the prior art can essentially be embodied in the form of computer software products. These computer software products can be stored in computer-readable storage media, such as ROM, RAM, disks, or optical discs, and contain several instructions to cause a computer device to execute the morel mushroom drying state image analysis, feature extraction, and modeling prediction methods described in the embodiments of the present invention or in parts thereof. The computer device can be a personal computer, server, or other network device.

[0061] Example 1: 1) Sample acquisition and imaging condition setting Fresh morel mushrooms were selected as the research object and dried in stages under hot air drying conditions. The drying temperature was controlled within the range of 45–60 °C. Samples were collected at different drying stages at fixed time intervals. Visible light RGB imaging system was used for image acquisition. The camera resolution was not less than 2000×2000 px, and the vertical distance between the lens and the sample surface was controlled at 30–50 cm. The camera parameters were kept consistent throughout the experiment to ensure the comparability of images in terms of spatial scale and brightness distribution.

[0062] 2) Image preprocessing and cap region construction The acquired raw RGB images were scaled uniformly, with the long side of the image scaled to 1024 px. Then, brightness normalization and background suppression were performed. Non-target background areas were removed by combining grayscale thresholding and morphological opening and closing operations. Based on this, the largest connected region was extracted through connected component analysis to construct a region of interest (ROI) containing only the cap region for subsequent feature analysis.

[0063] 3) Reinforced hole boundaries and prominent local structures To address the issue of blurred boundaries between pores on the surface of mushroom caps, edge enhancement processing is performed on the ROI images of mushroom caps. After suppressing high-frequency noise by using Gaussian smoothing (kernel size 15–31 px), a gradient operator is introduced to enhance the pore boundaries, thereby improving the contrast and distinguishability of the pore outline and internal structural features.

[0064] 4) Hole instance-level recognition and segmentation Based on the enhanced cap images, a deep learning-based instance segmentation model is used to identify holes, and each hole is labeled as an independent instance. For high-density areas, multi-scale reasoning and slice reasoning strategies are introduced to reduce the risk of hole adhesion and missed detection, and finally obtain complete cap hole instance-level segmentation results.

[0065] 5) Quantitative calculation of the structural characteristics of pores Based on the instance-level segmentation results, the hole area (number of pixels and its proportion of the cap area), equivalent diameter, boundary perimeter and boundary complexity index are calculated for each hole instance; at the same time, the hole number density, the nearest neighbor distance between holes and the proportion of hole connected regions are statistically analyzed to form a systematic set of hole structure features.

[0066] 6) Extraction of overall morphological features of the cap On the same ROI of the cap, the overall morphology of the cap is analyzed, and geometric parameters such as the cap's projected area, overall perimeter, principal axis length, secondary axis length, and aspect ratio are calculated to reflect the overall shrinkage, deformation, and morphological rearrangement characteristics of the cap during the drying process.

[0067] 7) Color Feature Extraction Convert the mushroom cap ROI image to the CIE-Lab color space and extract the brightness parameters. and chromaticity parameters The mean, standard deviation, and other statistical measures are used to characterize the changes in the brightness of the cap surface and the browning trend during the drying process.

[0068] 8) Multi-feature fusion and feature vector construction The pore structure, morphological features, and color features are standardized and then spliced ​​in a predetermined order to construct a unified multidimensional feature vector, which comprehensively reflects the phenotypic structural evolution of morel caps during the drying process.

[0069] 9) Regression Model Establishment and Dryness Prediction Based on the multidimensional feature vector, a regression model is selected to model the mapping relationship between features and drying state indicators. The quantitative relationship between pore structure, morphology and color features and drying degree is learned through training samples. After the model is trained, it is applied to test samples to output predicted values. The model performance is evaluated by comparing the predicted results with the measured values, thereby realizing the quantitative monitoring and state characterization of the morel drying process.

[0070] In this embodiment, a gradient boosting-based regression model is preferably used to model the drying state of morel mushrooms. After the model is trained, the comprehensive feature vector of the test samples is input into the regression model to obtain the corresponding predicted drying state values. Furthermore, the model prediction results are compared with the actually measured drying state indicators, and visualized in the form of a scatter plot. The horizontal axis represents the actual drying state value, the vertical axis represents the model prediction value, and the dashed line represents the reference relationship where the predicted value and the actual value are equal under ideal prediction conditions. The prediction effect is as follows: Figure 9 As shown in the figure, the prediction results show that the model's predicted values ​​are in good agreement with the actual values, indicating that the regression model established based on the fusion of pore structure features, morphological features, and color features can effectively reflect the state changes of morel mushrooms during the drying process and achieve stable prediction of the degree of drying.

[0071] Example 2: A method for predicting the drying state of morel mushrooms based on original features and a linear regression model Based on the morel mushroom image acquisition, preprocessing, and feature extraction process described in Example 1, this example directly uses the extracted raw features as model input without differentiating or standardizing. Lasso regression, Ridge regression, and ElasticNet regression models are constructed to quantitatively predict the drying state of morel mushrooms. By comparing the predicted results with the actual water loss rate data, the mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R²) of the models are calculated to evaluate the predictive performance of different linear models under the original feature conditions.

[0072] Example 3: A method for predicting the drying state of morel mushrooms based on baseline difference features and ensemble learning models Building upon the feature extraction described in Example 1, this example further introduces a baseline difference feature construction method. This involves using the initial feature value as a benchmark and performing difference processing on the corresponding features at each time point to form baseline difference features (X–X0). Based on this, random forest regression models, gradient boosting regression models, and XGBoost regression models are constructed to predict the drying state of morel mushrooms, and the prediction results of each model are quantitatively evaluated.

[0073] Example 4: A method for predicting the drying state of morel mushrooms based on standardized feature processing After the original feature extraction described in Example 1 is completed, this example performs standardization on the feature data to eliminate the impact of differences in feature dimensions on the model training process. Under the standardized feature conditions, Lasso, Ridge, and ElasticNet regression models are constructed respectively to predict the drying state of morel mushrooms, and the results are compared with those of the model without standardization to evaluate the impact of standardization on prediction performance.

[0074] Example 5: A method for predicting the drying state of morel mushrooms based on feature screening preprocessing After feature extraction as described in Example 1, this example further introduces a feature selection and preprocessing method based on Lasso regression and Ridge regression to compress and select high-dimensional features, thereby reducing feature redundancy. After feature selection, a corresponding regression model is constructed based on the selected features, and the drying state of morel mushrooms is predicted to verify the effect of the feature selection strategy in improving the generalization ability of the model.

[0075] A comprehensive comparative analysis of different feature processing methods and regression model combinations Based on Examples 2 to 5, this example summarizes and compares the prediction results under different feature processing methods and regression model combinations. The MAE, RMSE, and R² indices of each model in the morel drying state prediction task are shown in Appendix Table 1. The results show that the prediction scheme based on baseline difference features combined with the XGBoost regression model performs best in terms of accuracy and stability, further verifying the effectiveness and practical value of the multi-feature fusion and modeling method proposed in this invention for predicting the morel drying state.

[0076] Appendix 1

[0077] Example 2 This invention proposes a visual recognition and state characterization system for morel mushroom cap pores during the drying process, such as... Figure 10 As shown, it includes: A hole segmentation module is used to construct the region of interest of morel mushroom caps, and based on the region of interest of morel mushroom caps, to perform instance-level identification and segmentation of individual holes on the surface of the caps to obtain hole segmentation results. The feature extraction module is used to extract hole structure features based on hole segmentation results, and extract mushroom cap morphology features and mushroom cap color features based on the region of interest of the mushroom cap. The feature fusion module is used to fuse the features of the pore structure, the morphology of the mushroom cap, and the color of the mushroom cap to construct a multi-dimensional feature vector. The result prediction module is used to input multidimensional feature vectors into a pre-trained morel drying state prediction regression model and output the morel drying state characterization results to realize the monitoring of the morel drying process.

[0078] Example 3 Please see Figure 11 As shown, the present invention also provides an electronic device 100 for a method of visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0079] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the method for visual recognition and state characterization of cap pores in the morel drying process monitoring described in Embodiment 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device 100 (such as audio data), etc. In addition, the memory 101 may include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other non-volatile solid-state storage device.

[0080] The at least one processor 102 may be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor. The processor 102 is the control center of the electronic device 100, connecting various parts of the electronic device 100 via various interfaces and lines.

[0081] The memory 101 in the electronic device 100 stores multiple instructions to implement a method for visual recognition and state characterization of cap pores during the morel drying process, and the processor 102 can execute the multiple instructions to achieve the following: Construct the region of interest (ROI) of the morel mushroom cap, and based on the ROI, perform instance-level identification and segmentation of individual pores on the cap surface to obtain pore segmentation results; Based on the hole segmentation results, hole structure features are extracted, and mushroom cap morphology and color features are extracted based on the region of interest of the mushroom cap. The features of pore structure, cap morphology and color are fused to construct a multi-dimensional feature vector; By inputting multidimensional feature vectors into a pre-trained morel drying state prediction regression model, the model outputs the morel drying state characterization results, thereby enabling the monitoring of the morel drying process.

[0082] Example 4 If the modules / units integrated in the electronic device 100 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, and a read-only memory (ROM).

[0083] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0084] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0085] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for visual identification and state characterization of cap pores during the drying process of morel mushrooms, characterized in that, Includes the following steps: Construct the region of interest (ROI) of the morel mushroom cap, and based on the ROI, perform instance-level identification and segmentation of individual pores on the cap surface to obtain pore segmentation results; Based on the hole segmentation results, hole structure features are extracted, and mushroom cap morphology and color features are extracted based on the region of interest of the mushroom cap. The features of pore structure, cap morphology and color are fused to construct a multi-dimensional feature vector; By inputting multidimensional feature vectors into a pre-trained morel drying state prediction regression model, the model outputs the morel drying state characterization results, thereby enabling the monitoring of the morel drying process.

2. The method for visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms according to claim 1, characterized in that, The construction of the region of interest for the morel mushroom cap specifically involves: The original cap surface image is converted to grayscale to obtain a grayscale image of the cap surface. The grayscale image is then inverted and subjected to strong Gaussian smoothing. After smoothing, an adaptive threshold segmentation method is used for binarization to obtain an initial binary image of the cap region. Morphological closing operation is performed on the binary image to obtain a mask image of the cap region, thus constructing the region of interest of the morel cap.

3. The method for visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms according to claim 1, characterized in that, The process involves instance-level identification and segmentation of individual pores on the surface of the morel cap, based on the region of interest of the cap, to obtain the pore segmentation results. Specifically: The region of interest of the morel mushroom cap is divided into several overlapping or non-overlapping local sub-regions according to a preset size. The hole structure is independently identified in each sub-region. During the identification process, each identified hole instance is independently marked and a corresponding instance segmentation mask is generated to obtain the hole instance-level segmentation result.

4. The method for visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms according to claim 1, characterized in that, The extraction of hole structure features based on hole segmentation results specifically includes: The pore structure features include pore area, the ratio of pore area to the projected area of ​​the cap, the number of pores, the distribution of equivalent diameter of pores, the spatial distribution density between pores, and the connectivity features of pores. By comparing and analyzing the pore structure characteristics at different drying stages, the evolution trend of pores from the initial state to the shrinkage, deformation and rearrangement state during the drying process is quantitatively described.

5. The method for visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms according to claim 1, characterized in that, Based on the region of interest of the cap, the morphological features of the cap are extracted, specifically: The outline boundary of the mushroom cap is extracted, and the projected area and perimeter of the mushroom cap are calculated. The principal axis and secondary axis direction of the mushroom cap outline are analyzed, and the principal axis length and secondary axis length of the mushroom cap are calculated respectively. By the change relationship between the principal axis and secondary axis length, the anisotropic shrinkage behavior of the mushroom cap during the drying process is characterized.

6. The method for visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms according to claim 1, characterized in that, Based on the region of interest of the mushroom cap, the color features of the mushroom cap are extracted, specifically: Based on the region of interest (ROI) of the mushroom cap, statistical analysis was performed on the pixel color information on the cap surface, and the color space was converted from RGB to CIE. Color space; specifically, within the region of interest of the mushroom cap, brightness parameters are extracted pixel by pixel. and chromaticity parameters Furthermore, the color parameters under the same drying stage were statistically summarized to obtain the color distribution characteristics of the cap surface at that stage.

7. The method for visual identification and state characterization of cap pores for monitoring the drying process of morel mushrooms according to claim 1, characterized in that, The pre-trained regression model for predicting the drying state of morel mushrooms is as follows: Morel mushroom samples at different drying stages were used as training samples. The mapping relationship between the comprehensive feature vector and the corresponding drying state was learned through regression modeling. After the model training was completed, the trained regression model was applied to the test samples to predict the drying state of morel mushrooms and obtain the corresponding prediction results. The prediction results were compared and analyzed with the actual drying state and displayed in a visualization manner to evaluate the prediction performance of the model.

8. A system for visual recognition and state characterization of cap pores during the drying process of morel mushrooms, characterized in that, include: A hole segmentation module is used to construct the region of interest of morel mushroom caps, and based on the region of interest of morel mushroom caps, to perform instance-level identification and segmentation of individual holes on the surface of the caps to obtain hole segmentation results. The feature extraction module is used to extract hole structure features based on hole segmentation results, and extract mushroom cap morphology features and mushroom cap color features based on the region of interest of the mushroom cap. The feature fusion module is used to fuse the features of the pore structure, the morphology of the mushroom cap, and the color of the mushroom cap to construct a multi-dimensional feature vector. The result prediction module is used to input multidimensional feature vectors into a pre-trained morel drying state prediction regression model and output the morel drying state characterization results to realize the monitoring of the morel drying process.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for visual recognition and state characterization of cap pores for monitoring the morel drying process as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for visual recognition and state characterization of cap pores for monitoring the morel drying process as described in any one of claims 1 to 7.

Images