Method for identifying growth state of scirpus mariqueter under environmental stress by fusing texture color

Abstract

The invention relates to a method for identifying the growth state of scirpus mariqueter under environmental stress by fusing texture colors. The method comprises the following steps: step 1, sampling; step 2, acquiring an image; step 3, extracting RGB color features; step 4, extracting texture feature parameters; 5, key combination feature indexes are calculated; step 6, establishing a standard; step 7, verifying; and step 8, application. The method for identifying the growth state of scirpus mariqueter under environmental stress by fusing texture and color has the following advantages: early and tiny stress characterization which is difficult to perceive by a traditional naked eye observation method can be captured by quantitatively analyzing surface texture features and color features of leaf microscopic images and constructing key combination indexes; the whole process can be completed within several minutes; the equipment is simple, the cost is low, and the sampling strategy is optimized; and universality is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of plant physiological and ecological monitoring, and in particular to a method for identifying the growth state of Tripterygium wilfordii under environmental stress by fusing texture and color. Background Art

[0002] Bolboschoenoplectus mariqueter plays a crucial ecological role in coastal ecosystems. As a pioneer salt marsh species, it effectively mitigates coastal erosion, promotes sediment deposition, purifies water quality, sequesters carbon dioxide, and reduces emissions, thereby maintaining the stability of coastal ecosystems. Furthermore, Bolboschoenoplectus mariqueter provides habitat and food for a variety of birds, fish, and other wetland species, making it a key species in maintaining coastal wetland biodiversity.

[0003] In recent years, with the intensification of global climate change and human activities, the coastal wetlands where Scirpus maritima depends for survival have been continuously disturbed by multiple environmental stressors, including salinity fluctuations, water level changes, and heavy metal pollution. These stresses directly affect the growth of individual Scirpus maritima plants, their population dynamics, and the proper functioning of their ecological functions. Therefore, accurately monitoring and assessing the environmental stresses and growth status of Scirpus maritima has become a critical technical issue that needs to be addressed in coastal ecological conservation.

[0004] The traditional methods for identifying the growth status of Tripterygium wilfordii under environmental stress are visual observation and physiological and biochemical analysis. Visual observation cannot timely observe subtle stress signs when Tripterygium wilfordii is under environmental stress in the early stages. Physiological and biochemical analysis often requires a large sample size, which may cause great damage to the plants. In addition, although physiological and biochemical analysis can provide accurate information on the physiological status of plants, it usually involves complex sample processing and laboratory analysis processes, which are time-consuming (hours to days) and costly. Traditional identification methods have an irreconcilable contradiction between high precision and low cost / ease of operation. There is a lack of a practical solution that can guarantee high identification accuracy and reliability, while being cost-effective, easy to operate, and minimally invasive, and can adapt to the needs of large-scale and regular monitoring of coastal wetlands.

[0005] Therefore, a method of integrating texture and color to identify the growth status of Tripterygium wilfordii under environmental stress was designed to solve the above problems. Summary of the Invention

[0006] To overcome the above shortcomings, the present invention provides a method for identifying the growth state of Trichoderma marinum under environmental stress by fusing texture and color, comprising the following steps:

[0007] Step 1: Sampling:

[0008] Obtaining a leaf of Scirpus maritima and performing required processing on the leaf of Scirpus maritima before collecting a microscope image to obtain a sample of the leaf of Scirpus maritima;

[0009] Step 2: Capture images:

[0010] Using a stereo microscope to collect a microscopic image of the sample of the leaf of the Tripterygium wilfordii in step 1, and processing the microscopic image to generate first data;

[0011] Step 3: Extract RGB color features:

[0012] Extracting RGB color features from the first data to obtain second data;

[0013] Step 4: Extract texture feature parameters:

[0014] grayscale the first data to obtain third data; calculate a gray level co-occurrence matrix (GLCM) based on the third data, and extract texture feature parameters to obtain fourth data;

[0015] Step 5: Calculate key combination characteristic indicators:

[0016] Perform feature fusion on the second data obtained in step 3 and the fourth data obtained in step 4 to calculate a key combination feature index;

[0017] Step 6: Establish standards:

[0018] Establishing a growth state identification standard of Tripterygium wilfordii under salinity stress according to the selected key combination characteristic indicators;

[0019] Step 7: Verification:

[0020] To verify the accuracy and application effectiveness of the established identification standard for the growth status of Tripterygium wilfordii under environmental stress;

[0021] Step 8: Application:

[0022] According to the complete set of operating procedures disclosed and defined in steps 1 to 5, the three key combined indicators of each leaf sample are calculated;

[0023] The combined index value of each leaf sample is compared with the threshold range defined in step 6, and the growth status of each leaf sample is identified based on the comprehensive judgment rules.

[0024] Preferably, the processing method before microscope image acquisition in step 1 comprises the following steps:

[0025] The first step was to select representative Trichoderma seagrass plants from each salinity treatment group and cut leaves 1 cm from the tip of the leaves;

[0026] Step 2: Place the cut leaves in moist filter paper, put them in a sealed fresh-keeping box, and maintain the ambient temperature between 4-10°C;

[0027] Step 3: Use deionized water to gently clean the leaf surface to remove attached salt and impurities to obtain leaf samples;

[0028] Preferably, in step 2, the microscopic image is collected by:

[0029] Use a stereo microscope set to a fixed magnification, preferably 150x;

[0030] Adjust the microscope lighting system, set the light source brightness to 60-75% of the maximum brightness, and the light source angle to 45°±5°;

[0031] A color digital camera with a resolution of no less than 8 million pixels is installed on the microscope;

[0032] Spread the leaf sample flat on the stage to ensure that the surface of the leaf sample is flat and wrinkle-free. For each leaf sample, select the middle area of the leaf as the standard collection area.

[0033] Preferably, in step 2, the microscopic image is processed by:

[0034] Crop the images and adjust all images to a uniform size, preferably 500×500 pixels;

[0035] Perform image quality assessment to exclude images with substandard quality, such as those that are out of focus, underexposed or overexposed, or have obvious motion blur;

[0036] Perform color correction on the image and adjust the white balance using a standard color chart.

[0037] Preferably, in step 3, the RGB color feature is extracted by:

[0038] Extract the pixel values of the red (R), green (G), and blue (B) channels from the first data in step 1, and calculate the mean of each channel;

[0039] Based on the channel means, calculating a color ratio feature, preferably including an R / G ratio;

[0040] The normalized green-red difference index (NGRDI) was calculated to assess leaf health and yellowing degree using the following formula: NGRDI = (GR) / (G+R).

[0041] Preferably, in step 4, the grayscale processing method is:

[0042] Convert the preprocessed image into an 8-bit grayscale image. The conversion method uses the weighted average method, and the specific weight coefficients are: the weight of the red channel is 0.299, the weight of the green channel is 0.587, and the weight of the blue channel is 0.114.

[0043] Preferably, in the fourth step, the calculation method of the gray-level co-occurrence matrix (GLCM) is: set the offset parameter to 20; set the quantization level to 8 bits; set the calculation angle to 0°, that is, analyze the texture information in the horizontal direction.

[0044] Preferably, in the fourth step, the extraction method of the texture feature parameters is: based on the calculated gray-level co-occurrence matrix, extract the texture feature parameters of contrast, energy, and entropy.

[0045] Preferably, in the fifth step, the key combined feature indicators calculated are:

[0046] The Contrast*R / G ratio, which reflects the degree of coordination between "color deterioration" and "surface roughening";

[0047] The NGRDI*Energy value, which reflects the coupling state between "physiological health based on color" and "texture structure uniformity";

[0048] The Contrast*Entropy value, which reflects the "complexity" and "chaos" of the leaf surface texture.

[0049] Preferably, in the sixth step, the method for establishing the recognition standard is:

[0050] Use one-way analysis of variance (ANOVA) and combine it with multiple comparison methods to test whether there are statistically significant differences in the means of the three key combined indicators between different growth state categories;

[0051] According to the statistical analysis results, determine the combined indicator thresholds that can effectively divide the three state categories, specifically:

[0052] "Good" category (healthy state): contrast*R / G value < 0.09 and NGRDI*energy value > 0.13 and contrast*entropy value < 0.09;

[0053] "Medium" category (mild stress state): 0.09 ≤ contrast*R / G value < 0.16 and 0.05 < NGRDI*energy value ≤ 0.13 and 0.09 ≤ contrast*entropy value < 0.15;

[0054] "Poor" category (severe stress state): contrast*R / G value ≥ 0.16 and NGRDI*energy value ≤ 0.05 and contrast*entropy value ≥ 0.15.

[0055] Preferably, the step 6 further includes the following comprehensive judgment rules:

[0056] Majority rule: For a certain Scutellaria marinum sample, if the values of at least two combined indicators fall within the threshold range corresponding to the same growth state category, the sample is determined to belong to that growth state category;

[0057] Priority principle: If the values of the three key combination indicators fall within the threshold ranges of "good", "medium" and "poor" respectively, the growth status category indicated by the contrast*R / G value will be used as the final judgment result of the sample.

[0058] Preferably, in step seven, the verification process is: collecting independent verification samples, and calculating the three key combination indicators of each verification sample according to the full set of operating procedures disclosed and defined in steps one to five; comparing the combination indicator value of each verification sample with the threshold range defined in step S6, and predicting the growth status category of each verification sample based on the comprehensive judgment rules; comparing the prediction results with the benchmark growth status category to evaluate the accuracy of the recognition standard.

[0059] The beneficial effects of the present invention are:

[0060] The method of identifying the growth state of Tripterygium wilfordii under environmental stress by fusing texture and color according to the present invention has the following advantages:

[0061] 1. The present invention quantitatively analyzes the surface texture characteristics (such as contrast, energy, entropy, etc.) and color characteristics (such as R / G ratio, NGRDI, etc.) of leaf microscopic images, and constructs key combined indicators (contrast*R / G value, NGRDI*energy value and contrast*entropy value). It can capture early and subtle stress manifestations that are difficult to detect with the traditional naked eye.

[0062] 2. The present method relies on image acquisition and computational analysis, completing the entire process from sample acquisition to output of growth status determination results within minutes. This increases detection speed by dozens of times or even more, enabling high-frequency, near-real-time monitoring of the growth status of Tripterygium maritima.

[0063] 3. This method primarily relies on a stereo microscope and a conventional computer for image acquisition and data processing. The equipment required is relatively simple, and the cost is far lower than that of a fully equipped physiological and biochemical laboratory. The standardized operating procedures require relatively low operator expertise, and no complex chemical reagents or consumables are required. This not only significantly reduces the cost of a single test but also greatly lowers the threshold for technology application, facilitating large-scale, extensive ecological monitoring under resource-limited conditions.

[0064] 4. This invention optimizes the sampling strategy. By taking micro-samples only from the tip of the leaf, which is sensitive to stress reactions (for example, 1 cm from the tip), it can obtain key information reflecting the overall stress state of the plant while minimizing damage to the plant itself. This achieves a good balance between information acquisition and ecological protection, making it particularly suitable for monitoring rare or vulnerable plant populations.

[0065] 5. The plant stress status identification methodology established in this invention, based on the collaborative analysis of image texture and color features, is universally applicable. While this example uses Trichoderma maritima as an example, the method's technical framework and core indicators, after adaptive adjustments, are expected to be applied to the health status monitoring and stress assessment of other coastal wetland plants and even a wider range of vegetation types, providing a new and efficient, accurate, and economically viable technology for regional ecosystem protection and management. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a schematic diagram of the steps of the method for identifying the growth status of Trichoderma seagrass under environmental stress by fusing texture and color according to the present invention.

[0067] Figure 2 These are microscopic images and grayscale images of leaf samples of Tribolium maritima under different salinity stresses, collected using a stereo microscope in an embodiment of the present invention.

[0068] Figure 3 Schematic diagram of data distribution and multiple comparison significance differences of three key combination indicators under different salinity stresses in the embodiment of the present invention.

[0069] *Note: Different lowercase letters (a, b, c) indicate significant differences among different salinity treatment groups. DETAILED DESCRIPTION

[0070] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0071] See also Figures 2 to 3 The present invention provides a method for identifying the growth state of Trichoderma maritima under environmental stress (low, medium, and high salinity) based on surface texture and color features. The technical solution is as follows:

[0072] Example 1

[0073] In this example, a controlled indoor experiment was conducted to investigate the growth patterns of Tripterygium maritima under different salinity stress conditions. Six different salinity gradients (2‰, 6‰, 10‰, 14‰, 16‰, and 20‰) were set up in the laboratory. Three seedlings (13±2 cm) were placed in each salinity gradient treatment group for 14 days. During the experiment, a stable salinity level was maintained while recording the growth parameters and morphological changes of Tripterygium maritima.

[0074] Figure 1 The microscopic phenotype and grayscale images of the leaves of Tripterygium wilfordii under different salinity stresses provided in the embodiments of the present invention. As the degree of salinity stress increases, the surface texture and color of the leaves of Tripterygium wilfordii change significantly. The leaves under low salinity treatment (2‰-6‰) are uniformly green with regular texture, corresponding to a "good" growth state; the leaves under medium salinity treatment (10‰-14‰) begin to yellow and the texture becomes irregular, corresponding to a "medium" growth state; the leaves under high salinity treatment (16‰-20‰) are obviously yellow-brown with highly irregular texture, corresponding to a "poor" growth state.

[0075] like Figure 2 As shown, a method for identifying the growth state of Tripterygium wilfordii under environmental stress based on surface texture and color features comprises:

[0076] Step S1: obtaining a leaf sample of Tripterygium wilfordii, and collecting and preparing the leaf sample.

[0077] Furthermore, the leaf sample collection and preparation process includes:

[0078] From each salinity treatment group (2‰, 6‰, 10‰, 14‰, 16‰, 20‰), representative Marinus sclerotiorum plants were selected for growth status. Leaf samples were cut 1 cm from the leaf tip using sharp scissors. At least 3 samples were collected from each treatment group.

[0079] Place the collected leaf samples in moistened filter paper, store in a sealed fresh-keeping box, maintain the temperature between 4-10°C, and ensure that the samples are processed within 2 hours.

[0080] Use deionized water to gently clean the leaf surface to remove attached salt and impurities, then use absorbent paper to gently dry the surface moisture to avoid damaging the leaf surface tissue;

[0081] When there are no water droplets or obvious contaminants on the surface of the leaf, lay it flat on the stage to ensure that the surface of the leaf is flat and wrinkle-free.

[0082] Specifically, in this example, three plants of Trichoderma seaweed were selected from each of the six salinity gradient treatment groups. The tip of one leaf was collected from each plant, for a total of 18 leaf samples. These samples were initially categorized as "good" (primarily from the 2‰ and 6‰ treatment groups), "medium" (primarily from the 10‰ and 14‰ treatment groups), and "poor" (primarily from the 16‰ and 20‰ treatment groups) based on salinity treatment level and physiological performance.

[0083] Step S2: using a stereo microscope to collect a microscopic image of the leaf sample, and preprocessing the microscopic image to generate first data.

[0084] Furthermore, the microscopic image acquisition process includes:

[0085] Use a stereo microscope to set a fixed magnification, preferably 150 times, to ensure that the texture features of the leaf surface are clearly visible at this magnification;

[0086] Adjust the microscope lighting system, set the light source brightness to 60-75% of the maximum brightness and the light source angle to 45°±5° to produce uniform and sufficient lighting conditions to ensure clear texture details and color information in the image;

[0087] Install a color digital camera with a resolution of at least 8 megapixels on the microscope, set the ISO value to 100-200, the exposure time to 1 / 60-1 / 100 second, and the aperture value to F8-F11, and ensure that the acquired image is clear and the exposure is appropriate;

[0088] For each leaf sample, the middle area of the leaf was selected as the standard acquisition area, and at least three high-quality images were obtained for each sample;

[0089] Save the captured images in JPEG format with a resolution of no less than 2048 × 1536 pixels. The file naming convention is "salinity treatment level_sample number_serial number.jpg" to facilitate subsequent processing and analysis.

[0090] Furthermore, the preprocessing process of the microscopic image includes:

[0091] Crop the images and resize all images to a uniform size, preferably 500 × 500 pixels, to ensure that only the parts of the photos that are in focus are selected to facilitate consistency in subsequent analysis;

[0092] Perform image quality assessment to exclude images with unacceptable quality, such as those that are out of focus, underexposed or overexposed, or have obvious motion blur;

[0093] Perform color correction on the image and adjust the white balance using a standard color card to ensure the accuracy of color information;

[0094] The processed image is stored in JPEG format. The file name remains the same as the original image, but the "_processed" suffix is added to distinguish it.

[0095] Step S3: extracting RGB color features from the second data to obtain second data; and performing grayscale processing on the second data to obtain third data.

[0096] Furthermore, the RGB color feature extraction process includes:

[0097] Extract the pixel values of the red (R), green (G), and blue (B) channels from the preprocessed original color image and calculate the mean of each channel;

[0098] Based on the channel means, calculating a color ratio feature, preferably including an R / G ratio;

[0099] The normalized green-red difference index (NGRDI) was calculated to assess leaf health and yellowing degree;

[0100] Specifically, in this embodiment, NGRDI is calculated using the following formula: NGRDI = (GR) / (G + R), where G is the mean value of the green channel and R is the mean value of the red channel. Higher values generally indicate healthier leaves; lower values indicate a higher degree of yellowing or browning.

[0101] Furthermore, the grayscale processing includes:

[0102] The preprocessed image is converted to an 8-bit grayscale image (i.e., the grayscale level is 256). The conversion method uses the weighted average method. The specific weight coefficients are: red channel weight 0.299, green channel weight 0.587, and blue channel weight 0.114;

[0103] The grayscale image is stored in JPEG format. The file name remains the same as the pre-processed image, but the "_gray" suffix is added to distinguish it.

[0104] Step S4: Calculate a gray level co-occurrence matrix (GLCM) based on the fourth data, and extract texture feature parameters to obtain fourth data.

[0105] Furthermore, the calculation process of the gray level co-occurrence matrix (GLCM) includes:

[0106] Set the Offset parameter to

[20] , which means that the pixel pairs are 2 pixels apart in the horizontal direction;

[0107] The quantization level is set to 8 bits (i.e. 256 gray levels);

[0108] The calculation angle is set to 0°, that is, the texture information is analyzed in the horizontal direction.

[0109] Furthermore, the process of extracting the texture feature parameters includes:

[0110] Based on the calculated gray-level co-occurrence matrix, the following texture feature parameters are extracted:

[0111] Contrast: reflects the intensity of local changes in the image;

[0112] Energy: reflects the uniformity of image grayscale distribution and texture coarseness;

[0113] Entropy: describes the amount of information contained in an image;

[0114] Specifically, in this embodiment, the calculation of the above texture features is implemented by MATLAB software, and the specific code includes:

[0115]

[0116]

[0117]

[0118] Step S5: performing feature fusion on the second data and the fourth data to calculate key combination feature indicators.

[0119] Furthermore, the combined index of color features and texture features is calculated as follows:

[0120] contrast*R / G ratio: reflects the degree of synergy between "color degradation" and "surface roughening";

[0121] NGRDI*energy value: reflects the coupling status of "color-based physiological health" and "texture structure uniformity";

[0122] Contrast*entropy value: reflects the "complexity" and "chaos" of the leaf surface texture;

[0123] Step S6: establishing a growth state identification standard of Scutellaria maritima under salinity stress based on the selected key combination characteristic indicators.

[0124] Furthermore, the identification standard establishment process includes:

[0125] To confirm the ability of the selected indices to distinguish between different growth states, statistical analysis methods were used to process the experimental data. Specifically, a one-way analysis of variance (ANOVA) was used, combined with multiple comparisons, to test whether the means of the three key combined indices (contrast*R / G value, NGRDI*energy value, and contrast*entropy value) were statistically significantly different across the three growth state categories (i.e., different salinity treatment groups). The quantitative analysis results, including the means and standard deviations of each indicator within each treatment group, as well as comparisons of significant differences across multiple comparisons, are summarized and presented in Table 1.

[0126] Table 1 Changes in leaf indexes of Tripterygium wilfordii under different salinity treatments (mean ± SD)

[0127] Salinity treatment (‰) contrast*R / G value NGRDI*energy value contrast*entropy value Growth status classification 2-6 (low salinity) <![CDATA[0.06±0.04 a ]]> <![CDATA[0.15±0.08 a ]]> <![CDATA[0.07±0.05 a ]]> Good (health) 10-14 (medium salinity) <![CDATA[0.12±0.11 a ]]> <![CDATA[0.11±0.12 a ]]> <![CDATA[0.11±0.12 a ]]> Medium (moderate duress) 16-20 (high salinity) <![CDATA[0.21±0.10 b ]]> <![CDATA[-0.01±0.03 b ]]> <![CDATA[0.20±0.09 a ]]> Poor (severe stress)

[0128] *Note: Different lowercase letters (a, b, c) in the same column indicate significant differences in the mean values among different salinity treatment groups.

[0129] Table 1 clearly demonstrates that the texture + color combination index used in this study is sensitive to salinity stress gradients and can effectively distinguish different growth states of S. maritima. The contrast*R / G value shows a significant upward trend as salinity levels increase from low (2‰-6‰) to medium (10‰-14‰) and finally to high (16‰-20‰). This reflects the synergistic effect of increased salt stress leading to leaf yellowing (increased R / G ratio) and enhanced surface texture irregularity (increased contrast). In contrast, the NGRDI*energy value shows a clear downward trend. Under "good" conditions, its average value remains high; when stress reaches "medium," it drops sharply; and under "poor" conditions, it reaches even lower negative values. This significant shift from positive to negative values strongly indicates the combined physiological changes of decreased chlorophyll content (decreased NGRDI) and disrupted leaf texture uniformity (decreased energy) as stress intensifies. The contrast*entropy value also shows an increasing trend with increasing salinity. The results showed that both the local variation intensity (Contrast) and the overall complexity or information content (Entropy) of leaf surface texture increased with the increase of salinity stress.

[0130] Furthermore, in order to more intuitively demonstrate the distinguishing effect of the indicators, Figure 3It shows the data distribution and statistical differences of three key combined indicators in the embodiments of the present invention under low, medium, and high salinity stress treatments. For each combined indicator, the data distributions of different salinity treatment groups show an obvious separation trend, and each key combined indicator can effectively distinguish different growth states by itself. Specifically, the values of contrast*entropy and contrast*R / G increase significantly with the increase of salinity, while the value of NGRDI*energy decreases significantly with the increase of salinity.

[0131] Furthermore, based on the significant differences of single indicators revealed by the above statistical analysis (as shown in Table 1 and Figure 3 shown), the present invention determines the combined indicator thresholds that can effectively divide these three state categories, specifically:

[0132] "Good" category (healthy state): contrast*R / G value < 0.09 and NGRDI*energy value > 0.13 and contrast*entropy value < 0.09;

[0133] "Medium" category (mild stress state): 0.09 ≤ contrast*R / G value < 0.16 and 0.05 < NGRDI*energy value ≤ 0.13 and 0.09 ≤ contrast*entropy value < 0.15;

[0134] "Poor" category (severe stress state): contrast*R / G value ≥ 0.16 and NGRDI*energy value ≤ 0.05 and contrast*entropy value ≥ 0.15;

[0135] Furthermore, to handle the situation where the actual sample indicator values may not exactly meet the threshold ranges of a certain category, the following comprehensive judgment rules are formulated:

[0136] Comprehensive judgment rule one (majority principle): For a certain Scirpus mariqueter sample, if the values of at least two combined indicators fall within the threshold ranges corresponding to the same growth state category ("good", "medium", or "poor"), then it is determined that the sample belongs to this growth state category.

[0137] Comprehensive judgment rule two (priority principle): If the values of the three key combined indicators exactly fall within the threshold ranges of the "good", "medium", and "poor" three different categories respectively (that is, a situation of one vote against one vote against one vote occurs), then a ruling is made according to the preset priority indicator. In this embodiment, the contrast*R / G value is designated as the priority judgment basis, that is, the growth state category indicated by the contrast*R / G value is used as the final judgment result of the sample.

[0138] Step S7: Verify the accuracy and application effectiveness of the established identification standard for the growth status of Scutellaria maritima under environmental stress.

[0139] Furthermore, the verification process specifically includes:

[0140] To provide objective validation, 30 new independent validation samples were cultured and collected across the entire salinity gradient described above, forming a separate validation set. For each sample in this validation set, a "true" or "baseline" growth status category (i.e., "good," "medium," or "poor") was pre-determined based on the known salinity treatment level it had experienced. This baseline category served as a reference standard for subsequent accuracy assessments. In this validation set, 10 samples each fell into the "good," "medium," and "poor" baseline categories.

[0141] The 30 independent validation samples obtained strictly followed the full set of operational procedures disclosed and defined in steps S1 to S5 of the present invention. Three key combined indicators were calculated for each validation sample: contrast*R / G value, NGRDI*energy value, and contrast*entropy value, forming comprehensive characteristic data for the validation sample.

[0142] Furthermore, the comprehensive feature data (ie, the three key combination index values) calculated for each verification sample is input into the recognition standard established in step S6.

[0143] Specifically, the contrast*R / G value, NGRDI*energy value, and contrast*entropy value of each validation sample are compared with the threshold ranges for the three growth status categories of "good," "medium," and "poor" defined in step S6. The growth status category of each validation sample is predicted and determined strictly according to the comprehensive judgment rules (majority principle and priority principle) established in step S6. The growth status category ("good," "medium," or "poor") predicted for each validation sample by the method of the present invention (i.e., the identification criteria) is recorded.

[0144] Furthermore, the predicted growth status categories are compared one by one with the determined benchmark growth status categories to quantitatively evaluate the accuracy of the identification criteria. The specific quantitative evaluation results are as follows:

[0145] For the 10 validation samples whose benchmark category is “good” (healthy), the recognition accuracy of the method of the present invention is 90.0% (9 / 10).

[0146] For the 10 validation samples whose benchmark category was “medium” (mild stress), the recognition accuracy of the method of the present invention was 80.0% (8 / 10).

[0147] For the 10 validation samples whose benchmark category was “poor” (severe stress), the recognition accuracy of the method of the present invention was 90.0% (9 / 10).

[0148] The overall accuracy of the established identification criteria was 86.7% (26 / 30).

[0149] The verification results fully confirm that the growth status identification standard of Scutellaria baicalensis under environmental (salinity) stress based on the surface texture synergistic color characteristics (specifically the three key combination indicators of contrast*R / G value, NGRDI*energy value, and contrast*entropy value) of the present invention can be effectively applied to the rapid, accurate, low-cost, and minimally damaging evaluation of the growth status of unknown samples, verifying the feasibility of the technical solution of the present invention and its significant practical application effectiveness.

[0150] In the description of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside", "outside", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present invention.

[0151] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0152] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for identifying the growth state of Tripterygium wilfordii under environmental stress by fusing texture and color, characterized in that: The steps include: Step 1: Sampling: Obtaining a leaf of Scirpus maritima and performing required processing on the leaf of Scirpus maritima before collecting a microscope image to obtain a sample of the leaf of Scirpus maritima; Step 2: Capture images: Using a stereo microscope to collect a microscopic image of the sample of the leaf of the Tripterygium wilfordii in step 1, and processing the microscopic image to generate first data; Step 3: Extract RGB color features: Extracting RGB color features from the first data to obtain second data; Step 4: Extract texture feature parameters: grayscale the first data to obtain third data; calculate a gray level co-occurrence matrix (GLCM) based on the third data, and extract texture feature parameters to obtain fourth data; Step 5: Calculate key combination characteristic indicators: Perform feature fusion on the second data obtained in step 3 and the fourth data obtained in step 4 to calculate a key combination feature index; Step 6: Establish standards: Establishing a growth state identification standard of Tripterygium wilfordii under salinity stress according to the selected key combination characteristic indicators; Step 7: Verification: To verify the accuracy and application effectiveness of the established identification standard for the growth status of Tripterygium wilfordii under environmental stress; Step 8: Application: According to the complete set of operating procedures disclosed and defined in steps 1 to 5, the three key combined indicators of each leaf sample are calculated; The combined index value of each leaf sample is compared with the threshold range defined in step 6, and the growth status of each leaf sample is identified based on the comprehensive judgment rules.

2. The method for identifying the growth state of Tripterygium wilfordii under environmental stress by fusing texture and color according to claim 1, characterized in that: The processing method before microscope image acquisition in step 1 includes the following steps: The first step was to select representative Trichoderma seagrass plants from each salinity treatment group and cut leaves 1 cm from the tip of the leaves; Step 2: Place the cut leaves in moist filter paper, put them in a sealed fresh-keeping box, and maintain the ambient temperature between 4-10°C; The third step is to use deionized water to gently clean the leaf surface to remove attached salt and impurities and obtain leaf samples.

3. The method for identifying the growth state of Tripterygium wilfordii under environmental stress by fusing texture and color according to claim 1, characterized in that: In step 2, the microscopic image is collected in the following manner: Use a stereo microscope set to a fixed magnification, preferably 150x; Adjust the microscope lighting system, set the light source brightness to 60-75% of the maximum brightness, and the light source angle to 45°±5°; A color digital camera with a resolution of no less than 8 million pixels is installed on the microscope; Spread the leaf sample flat on the stage to ensure that the surface of the leaf sample is flat and wrinkle-free. For each leaf sample, select the middle area of the leaf as the standard collection area; In step 2, the microscopic image is processed as follows: Crop the images and adjust all images to a uniform size, preferably 500×500 pixels; Perform image quality assessment to exclude images with substandard quality, such as those that are out of focus, underexposed or overexposed, or have obvious motion blur; Perform color correction on the image and adjust the white balance using a standard color chart.

4. The method for identifying the growth state of Tripterygium wilfordii under environmental stress by fusing texture and color according to claim 1, characterized in that: In step 3, the RGB color feature is extracted as follows: Extract the pixel values of the red (R), green (G), and blue (B) channels from the first data in step 1, and calculate the mean of each channel; Based on the channel means, calculate color ratio features, preferably including the R / G ratio; Calculate the Normalized Green-Red Difference Index (NGRDI) for evaluating leaf health status and yellowing degree, and its calculation formula is: NGRDI = (G - R) / (G + R).

5. The method for identifying the growth state of Tripterygium wilfordii under environmental stress by fusing texture and color according to claim 1, characterized in that: In the fourth step, the grayscale processing method is: Convert the preprocessed image into an 8-bit grayscale image, and the conversion method adopts the weighted average method. The specific weight coefficients are: the weight of the red channel is 0.299, the weight of the green channel is 0.587, and the weight of the blue channel is 0.114; In the fourth step, the calculation method of the Gray-Level Co-occurrence Matrix (GLCM) is: set the offset parameter to 20; the quantization level is set to 8 bits; the calculation angle is set to 0°, that is, analyze the texture information in the horizontal direction.

6. The method for identifying the growth state of Trichosanthis maritima under environmental stress by fusing texture and color according to claim 1, characterized in that: In the fourth step, the extraction method of texture feature parameters is: based on the calculated Gray-Level Co-occurrence Matrix, extract texture feature parameters of Contrast, Energy, and Entropy.

7. The method for identifying the growth state of Trichosanthis maritima under environmental stress by fusing texture and color according to claim 6, characterized in that: In the fifth step, the key combined feature indicators calculated are: Contrast * R / G ratio, reflecting the synergistic degree of "color deterioration" and "surface roughening"; NGRDI * Energy value, reflecting the coupling state of "physiological health based on color" and "texture structure uniformity"; Contrast * Entropy value, reflecting the "complexity" and "disorder" of the leaf surface texture.

8. The method for identifying the growth state of Trichosanthis maritima under environmental stress by fusing texture and color according to claim 7, characterized in that: In the sixth step, the establishment method of the recognition standard is: Adopt one-way analysis of variance (ANOVA) and combine it with multiple comparison methods to test whether there are statistically significant differences in the means of the three key combined indicators among different growth state categories; According to the statistical analysis results, determine the combined indicator thresholds that can effectively divide the three state categories, specifically: "Good" category (healthy state): contrast * R / G value < 0.09 and NGRDI * energy value > 0.13 and contrast * entropy value < 0.09; "Medium" category (mild stress state): 0.09 ≤ contrast * R / G value < 0.16 and 0.05 < NGRDI * energy value ≤ 0.13 and 0.09 ≤ contrast * entropy value < 0.15; "Poor" category (severe stress state): contrast * R / G value ≥ 0.16 and NGRDI * energy value ≤ 0.05 and contrast * entropy value ≥ 0.

15.

9. The method for identifying the growth state of Trichosanthis maritima under environmental stress by fusing texture and color according to claim 7, characterized in that: In the sixth step, the following comprehensive judgment rules are also included: Majority principle: For a sample of Scirpus mariqueter, if the values of at least two combined indicators fall within the threshold range corresponding to the same growth state category, then it is determined that the sample belongs to this growth state category; Priority principle: If the values of the three key combination indicators fall within the threshold ranges of "good", "medium" and "poor" respectively, the growth status category indicated by the contrast*R / G value will be used as the final judgment result of the sample.

10. The method for identifying the growth state of Trichosanthis maritima under environmental stress by fusing texture and color according to claim 1, characterized in that: The verification process of step seven is as follows: collecting independent verification samples, calculating the three key combination indicators of each verification sample according to the complete set of operating procedures disclosed and defined in steps one to five; comparing the combination indicator value of each verification sample with the threshold range defined in step S6, and predicting the growth status category of each verification sample based on the comprehensive judgment rule; comparing the prediction result with the benchmark growth status category to evaluate the accuracy of the recognition standard.

Images