Pumpkin germplasm salt tolerance grade evaluation method based on true leaf color space statistical analysis

Abstract

The invention discloses a pumpkin germplasm salt tolerance grade evaluation method based on true leaf color space statistical analysis. The method realizes rapid and convenient detection of the salt stress degree of the pumpkin seedlings. According to the method, a self-adaptive segmentation algorithm is provided for relatively completely extracting salt stress information of leaves by utilizing true leaf images of pumpkin seedlings with different salt damage degrees, a salt stress state index BSD is designed for representing the salt stress degrees of the leaves, and an optimal pumpkin seedling salt stress grade estimation model is established. Based on a crop phenotype technology and an image recognition algorithm, the method can adapt to different application scenes, salt stress of pumpkin seedlings is detected, compared with a traditional identification method, the efficiency and accuracy are greatly improved, and important support can be provided for agricultural production and scientific research.

Description

Technical Field

[0001] The invention belongs to the field of agricultural bioinformatics and agricultural automation, and particularly relates to a method for evaluating pumpkin germplasm salt tolerance grade by true leaf color space statistical analysis. Background Art

[0002] Pumpkin belongs to the genus Cucurbitaceae. The three main cultivated species are Chinese pumpkin, Indian pumpkin and American pumpkin. Because pumpkin has strong stress resistance, well-developed root system and wide adaptability, it is often used as the rootstock of melon crops such as watermelon, melon and cucumber. In recent years, the problem of secondary salinization of soil has been serious, which has greatly affected the yield and quality of crops. Breeding salt-resistant pumpkin varieties as rootstocks for grafting has become an effective method to improve the salt resistance of crops. At present, the methods for judging the degree of salt stress of pumpkin are mainly manual rating and laboratory physical and chemical analysis. However, these methods have high labor costs, poor timeliness and the results are easily affected by subjective judgment, low accuracy and poor repeatability. Therefore, there is an urgent need for a more scientific and convenient detection method to detect the degree of salt stress of pumpkin seedlings, so as to facilitate experimenters to better select and breed.

[0003] In recent years, the progress of related technologies such as machine vision and image processing has expanded the research on crop salt stress detection. Feng Xuping et al. used okra hyperspectral images to establish an inversion model of hyperspectral information and parameters such as SPAD value, sodium ion concentration, photosynthetic rate and transpiration rate, providing a reference for high-throughput phenotyping analysis of salt stress. Shuiqin Zhou et al. collected salt stress images of five soybean varieties through an image-based phenotyping system, and established a new characteristic parameter using the ratio of damaged leaf area to canopy area, realizing the quantitative rating of soybean salt tolerance. Jiayi Li et al. built a phenotyping platform for synchronous acquisition of multi-color fluorescence and multi-spectral reflectance images. By calculating the vegetation index, it was found that salt stress caused the green index (GR) to decrease, while the plant senescence reflectance index (PSRI) increased, providing important indicators for cotton salt stress diagnosis. Montanaro G. et al. collected images of grapevines treated with salt stress and soil conductivity, and designed a color rendering index CI based on RGB images, which can better predict soil conductivity and provide ideas for predicting salt stress in grapevines.

[0004] Although current research has provided many methods and references for salt stress detection, due to the complex leaf traits of pumpkin seedlings after salt stress, there is no scientific and accurate indicator for the image characterization of the salt stress state of pumpkin seedlings, and there is no report on the detection of salt stress level of pumpkin seedlings based on crop phenotyping platform. Summary of the invention

[0005] 1. Technical issues to be resolved In order to scientifically and conveniently identify the degree of salt stress of pumpkin seedlings, thereby facilitating experimenters to better carry out seed selection and breeding, the present invention provides a method for evaluating the salt tolerance level of pumpkin germplasm by statistical analysis of true leaf color space, which realizes the rapid and convenient detection of the degree of salt stress of pumpkin seedlings.

[0006] (II) Technical solution In order to solve the above problems, the present invention provides the following technical solutions, and proposes a method for evaluating the salt tolerance level of pumpkin germplasm based on true leaf color space statistical analysis, which is as follows.

[0007] A method for evaluating pumpkin germplasm salt tolerance grade based on true leaf color space statistical analysis comprises the following steps: Step S1, artificially grading the salt damage of pumpkin seedlings to obtain a salt damage index SI; Step S2, cutting off the true leaves of the pumpkin seedlings after the artificial salt damage rating, and collecting the RGB images of the true leaves of the pumpkin seedlings; Step S3, performing color correction on the collected RGB image; Step S4, cropping the color-corrected image, and taking a cell consisting of N pumpkin seedlings of the same variety as a region of interest ROI for image segmentation, where N is a positive integer, N≥8; Step S5, for the acquired ROI image, calculating the super green shape index EGSI, automatically finding the threshold through an adaptive threshold segmentation algorithm, and generating a pumpkin seedling binary mask image; Step S6, based on the pumpkin seedling binary mask image, obtaining the most probable value of the RGB three-channel intensity statistical result, and calculating the salt stress state index BSD based on the most probable value of the result; Step S7, performing inverse fitting on the salt stress state index BSD and the salt damage index SI to establish a pumpkin seedling salt stress level assessment model; Step S8, applying the pumpkin seedling salt stress level assessment model to the pumpkin germplasm salt tolerance experiment, specifically, first collecting the RGB image of the true leaves of the pumpkin seedlings, and then applying the pumpkin seedling salt stress level assessment model to assess the salt tolerance of different varieties of pumpkin seedlings.

[0008] Preferably, in step S1, when the pumpkin seedlings are graded for artificial salt damage, they are rated into 5 levels according to the salt tolerance standard, where level 0 means that the color of the first true leaf is normal, level 1 means that the edge of the first true leaf wilts and loses water, level 2 means that the edge of the first true leaf loses water and is scorched, level 3 means that the first true leaf loses water, wilts and is scorched, the plant loses water, wilts and falls, and level 4 means that the plant is dead; the salt damage index is calculated according to the following formula: .

[0009] Preferably, in step S2, a high-definition camera is used to obtain images of true leaves of pumpkin seedlings, and an image is collected once for every two rows of pumpkin seedlings.

[0010] Preferably, in step S3, a 24-color standard color card is used to correct the color changes caused by different lighting environments; the chromaticity values ​​extracted from the photographed 24-color standard color card are regressed and fitted with the chromaticity values ​​of the real colorimetric card to obtain the equation corresponding to each band, and the equation is substituted into each picture for calculation to correct the image.

[0011] Preferably, in step S5, the EGSI values ​​corresponding to all pixels in each ROI image obtained by segmentation are calculated, statistical analysis is performed to obtain a statistical graph with two peaks, and the critical positions of the two peaks are automatically selected as thresholds to generate a mask, separate the background from the leaves, and generate a pumpkin seedling binary mask image.

[0012] Preferably, in step S5, the calculation formula of the super green shape index EGSI is: Among them, R, G, and B correspond to the grayscale values ​​of the R, G, and B channels in the pumpkin seedling true leaf image, respectively.

[0013] Preferably, in step S6, based on the pumpkin seedling binary mask image, the three RGB channels are separated and normalized, and then the most probable values ​​of the RGB intensities are obtained respectively to calculate the salt stress state index BSD.

[0014] Preferably, the salt stress state index BSD calculation formula mentioned in step S6 is: in, , , They correspond to the most probable values ​​of the R, G, and B channel intensities respectively.

[0015] Preferably, in step S7, the salt damage index SI is set as the dependent variable, the salt stress state index BSD is used as the independent variable, 70% of the samples are selected for regression fitting, and a piecewise function is used for fitting.

[0016] Preferably, in step S8, the pumpkin seedling salt stress level assessment model is as follows: SI=-2.463*BSD+1.151, (BSD<0.35) SI=0.276, (BSD≥0.35) Wherein SI is the salt damage index obtained in step S1, and BSD is the salt stress state index calculated in step S6.

[0017] (III) Beneficial effects Compared with the prior art, the present invention has significant positive technical effects, which are specifically manifested in the following aspects.

[0018] (1) In the prior art, when using traditional spectral indices to generate masks for pumpkin seedling true leaf images, there is a problem of incomplete extraction of image salt stress feature information. However, the present application effectively extracts a relatively complete pumpkin seedling true leaf mask image by designing the Super Green Shape Index (EGSI), while retaining the complete salt stress feature information. This makes it possible to characterize the salt stress state of pumpkin seedlings by mining the correlation between color information statistics.

[0019] (2) In the prior art, there is no unified and objective evaluation standard for pumpkin salt tolerance evaluation. The accuracy and reliability of the evaluation results are low, making it difficult to accurately judge the degree of salt stress of the plants, which affects subsequent scientific research and planting decisions and increases the uncertainty of agricultural production. The present application effectively quantifies the degree of salt stress of pumpkin seedlings by designing a salt stress state index BSD, thereby achieving accurate and objective identification of the degree of salt stress of pumpkin seedlings.

[0020] (3) In the prior art, the identification of salt stress of pumpkin seedlings mainly relies on manual rating and laboratory physical and chemical analysis, which is time-consuming, labor-intensive, inefficient, and costly. In the context of large-scale planting and rapid development of modern agriculture, it cannot meet the needs of high-throughput plant salt tolerance rating. The present application designs a method for evaluating the salt tolerance level of pumpkin germplasm, specifically, batches of pumpkin seedling true leaf images are obtained, and segmentation and feature extraction are performed through an adaptive algorithm, so as to effectively and accurately rate the salt stress of pumpkin seedlings, thereby achieving a high-throughput pumpkin seedling salt stress level assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 The figure is a flow chart of the method of the present invention.

[0022] Figure 2 The invention discloses a pumpkin seedling salt stress grade assessment model.

[0023] Figure 3 Comparison of the adaptive segmentation algorithm and the classic spectral index segmentation results.

[0024] Figure 4 This is a graph showing the performance test results of the pumpkin seedling salt stress level assessment model used in the present invention.

[0025] Figure 5 This is a performance test result diagram of the pumpkin germplasm salt tolerance grade evaluation model used in the present invention. DETAILED DESCRIPTION

[0026] The present invention is further described below in conjunction with the accompanying drawings and implementation examples.

[0027] A method for evaluating the salt tolerance of pumpkin germplasm based on true leaf color space statistical analysis, the process is as follows Figure 1 As shown, the following steps are included: Step S1, artificially grading the salt damage of pumpkin seedlings to obtain a salt damage index SI; Step S2, cutting off the true leaves of the pumpkin seedlings after the artificial salt damage rating, and collecting the RGB images of the true leaves of the pumpkin seedlings; Step S3, performing color correction on the collected RGB image; Step S4, cropping the color-corrected image, and taking a cell consisting of N pumpkin seedlings of the same variety as a region of interest ROI for image segmentation, where N is a positive integer, N≥8; Step S5, for the acquired ROI image, calculating the super green shape index EGSI, automatically finding the threshold through an adaptive threshold segmentation algorithm, and generating a pumpkin seedling binary mask image; Step S6, based on the pumpkin seedling binary mask image, obtaining the most probable value of the RGB three-channel intensity statistical result, and calculating the salt stress state index BSD based on the most probable value of the result; Step S7, inverse fitting the salt stress state index BSD and the salt damage index SI is established as follows: Figure 2 Salt stress level assessment model for pumpkin seedlings; Step S8, applying the pumpkin seedling salt stress level assessment model to the pumpkin germplasm salt tolerance experiment, specifically, first collecting the RGB image of the true leaves of the pumpkin seedlings, and then applying the pumpkin seedling salt stress level assessment model to assess the salt tolerance of different varieties of pumpkin seedlings.

[0028] Preferably, in step S1, when the pumpkin seedlings are graded for artificial salt damage, they are rated into 5 levels according to the salt tolerance standard, where level 0 means that the color of the first true leaf is normal, level 1 means that the edge of the first true leaf wilts and loses water, level 2 means that the edge of the first true leaf loses water and is scorched, level 3 means that the first true leaf loses water, wilts and is scorched, the plant loses water, wilts and falls, and level 4 means that the plant is dead; the salt damage index SI is calculated according to the salt damage index formula: .

[0029] Preferably, in step S2, a high-definition camera is used to obtain images of true leaves of pumpkin seedlings, and an image is collected once for every two rows of pumpkin seedlings.

[0030] Preferably, in step S3, a 24-color standard color card is used to correct the color changes caused by different lighting environments; the chromaticity values ​​extracted from the photographed 24-color standard color card are regressed and fitted with the chromaticity values ​​of the real colorimetric card to obtain the equation corresponding to each band, and the equation is substituted into each picture for calculation to correct the image.

[0031] Preferably, in step S5, the EGSI values ​​corresponding to all pixels in each ROI image obtained by segmentation are calculated, statistical analysis is performed to obtain a statistical graph with two peaks, and the critical positions of the two peaks are automatically selected as thresholds to generate a mask, separate the background from the leaves, and generate a pumpkin seedling binary mask image.

[0032] Preferably, in step S5, the calculation formula of the super green shape index EGSI is: Among them, R, G, and B correspond to the grayscale values ​​of the R, G, and B channels in the pumpkin seedling true leaf image, respectively.

[0033] Preferably, in step S6, based on the pumpkin seedling binary mask image, the three RGB channels are separated and normalized, and then the most probable values ​​of the RGB intensities are obtained respectively to calculate the salt stress state index BSD.

[0034] Preferably, the salt stress state index BSD calculation formula mentioned in step S6 is: in, , , They correspond to the most probable values ​​of the R, G, and B channel intensities respectively.

[0035] Preferably, in step S7, the salt damage index SI is set as the dependent variable, the salt stress state index BSD is used as the independent variable, 70% of the samples are selected for regression fitting, and a piecewise function is used for fitting.

[0036] Preferably, in step S8, the pumpkin seedling salt stress level assessment model is as follows: SI=-2.463*BSD+1.151, (BSD<0.35) SI=0.276, (BSD≥0.35) Wherein SI is the salt damage index obtained in step S1, and BSD is the salt stress state index calculated in step S6.

[0037] In order to better explain the technical solution of the present invention, taking pumpkin seedlings as specific application objects, the present invention obtained the following results.

[0038] (1) The super green shape index (EGSI) is more effective than the traditional spectral index in extracting salt stress features. Due to the large differences in leaf structure, color and size, the binary mask obtained using the traditional spectral index is Figure 3 As shown in (c), the generated binary mask has a large error in the withering information, and some withered leaves are not removed, which makes it impossible to characterize the stress state by calculating the ratio of the withered area to the intact leaf area. Figure 3 (b) shows that the true leaves of pumpkin seedlings can be extracted relatively completely while retaining complete salt stress characteristic information.

[0039] (2) Evaluation effect of salt stress level of pumpkin seedlings. Based on the method proposed in the present invention, the true leaves of pumpkin seedlings were subjected to salt stress evaluation. Figure 4 As shown, the salt stress rating model R of pumpkin seedlings 2 The prediction accuracy is 84.79%, which can effectively detect the salt stress degree of pumpkin seedlings.

[0040] (3) Evaluation effect of pumpkin germplasm salt tolerance grade. Based on the method proposed in the present invention, 49 pumpkin varieties were selected, and the corresponding BSD mean values ​​were calculated. The pumpkin seedling salt stress rating model constructed using the BSD index was used to rate the salt tolerance of each pumpkin seedling material. The salt tolerance grades of each pumpkin variety were compared with the three-level salt tolerance grade classification standard (strong salt tolerance SI < 35%, medium 35% ≤ SI < 65%, weak SI ≥ 65%). The rating accuracy calculated using the formula shown below was 89.79%. The calculation results are as follows: Figure 5 It can be found that the present invention can more accurately rate the salt stress of pumpkin seedlings.

[0041] in, .

[0042] The specific examples described in the application are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific examples described in the present invention, or replace them in a similar manner, but they will not deviate from the spirit of the present invention or exceed the scope defined by the attached claims.

Claims

1. A method for evaluating the salt tolerance of pumpkin germplasm by statistical analysis of true leaf color space, characterized in that: The steps include: Step S1, artificially grading the salt damage of pumpkin seedlings to obtain a salt damage index SI; Step S2, cutting off the true leaves of the pumpkin seedlings after the artificial salt damage rating, and collecting the RGB images of the true leaves of the pumpkin seedlings; Step S3, performing color correction on the collected RGB image; Step S4, cropping the color-corrected image, and taking a cell consisting of N pumpkin seedlings of the same variety as a region of interest ROI for image segmentation, where N is a positive integer, N≥8; Step S5, for the acquired ROI image, calculating the super green shape index EGSI, automatically finding the threshold through an adaptive threshold segmentation algorithm, and generating a pumpkin seedling binary mask image; Step S6, based on the pumpkin seedling binary mask image, obtaining the most probable value of the RGB three-channel intensity statistical result, and calculating the salt stress state index BSD based on the most probable value of the result; Step S7, performing inverse fitting on the salt stress state index BSD and the salt damage index SI to establish a pumpkin seedling salt stress level assessment model; Step S8, applying the pumpkin seedling salt stress level assessment model to the pumpkin germplasm salt tolerance experiment, specifically, first collecting the RGB image of the true leaves of the pumpkin seedlings, and then applying the pumpkin seedling salt stress level assessment model to assess the salt tolerance of different varieties of pumpkin seedlings.

2. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 1, characterized in that: In step S1, when the pumpkin seedlings are graded for artificial salt damage, they are rated into 5 levels according to the salt tolerance standard, where level 0 means that the color of the first true leaf is normal, level 1 means that the edge of the first true leaf wilts and loses water, level 2 means that the edge of the first true leaf loses water and is scorched, level 3 means that the first true leaf loses water, wilts and scorches, the plant loses water, wilts and falls, and level 4 means that the plant is dead. The salt damage index is calculated according to the following formula: 。 3. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 1, characterized in that: In step S2, a high-definition camera is used to obtain images of true leaves of pumpkin seedlings, and an image is collected once for every two rows of pumpkin seedlings.

4. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 1, characterized in that: In step S3, a 24-color standard color card is used to correct the color changes caused by different lighting environments; the chromaticity values ​​extracted from the photographed 24-color standard color card are regressed and fitted with the chromaticity values ​​of the real colorimetric card to obtain the equation corresponding to each band, and the equation is substituted into each picture for calculation to correct the image.

5. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 1, characterized in that: In step S5, the EGSI values ​​corresponding to all pixels in each ROI image obtained by segmentation are calculated, and statistical analysis is performed to obtain a statistical graph with two peaks. The critical positions of the two peaks are automatically selected as thresholds to generate a mask, separate the background from the leaves, and generate a pumpkin seedling binary mask image.

6. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 5, characterized in that: In step S5, the calculation formula of the super green shape index EGSI is: Among them, R, G, and B correspond to the grayscale values ​​of the R, G, and B channels in the pumpkin seedling true leaf image, respectively.

7. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 1, characterized in that: In step S6, based on the pumpkin seedling binary mask image, the three RGB channels are separated and normalized, and then the most probable values ​​of RGB intensities are obtained respectively to calculate the salt stress state index BSD.

8. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 7, characterized in that: The calculation formula of the salt stress state index BSD mentioned in step S6 is: in, , , They correspond to the most probable values ​​of the R, G, and B channel intensities respectively.

9. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 1, characterized in that: In step S7, the salt damage index SI is set as the dependent variable, the salt stress state index BSD is used as the independent variable, 70% of the samples are selected for regression fitting, and a piecewise function is used for fitting.

10. The method for evaluating pumpkin germplasm salt tolerance by true leaf color space statistical analysis according to claim 1, characterized in that: In step S8, the pumpkin seedling salt stress level assessment model is as follows: SI=-2.463*BSD+1.151, (BSD<0.35) SI=0.276, (BSD≥0.35) Wherein SI is the salt damage index obtained in step S1, and BSD is the salt stress state index calculated in step S6.

Images