A phenotypic image method for rapidly screening sugarcane germplasm with different nitrogen use efficiency

Abstract

The application discloses a kind of phenotypic image methods for quickly screening sugarcane nitrogen utilization efficiency difference germplasm, it is the phenotypic two-dimensional image of different nitrogen utilization rate of sugarcane germplasm using LemnaTech 3D high-throughput phenotyping platform;After the image is collected, image cutting extraction is carried out by image processing software, to extract the phenotypic trait index related to sugarcane;After each data trait index is collected, it is calculated after standardization processing, and then data verification is carried out, and the key phenotypic index of sugarcane nitrogen utilization efficiency germplasm is screened;The regression model of key phenotypic index and sugarcane biomass and nitrogen utilization rate is established by multiple linear regression method.The key phenotypic trait index obtained by the method has a significant positive correlation with the nitrogen utilization rate and biomass of sugarcane, and can be used as a screening index for the difference of nitrogen utilization efficiency of sugarcane, to provide a theoretical reference for the growth of sugarcane and the breeding of nitrogen-efficient varieties.

Description

Technical Field

[0001] This invention relates to the field of agricultural technology, specifically a phenotypic image method for rapidly screening sugarcane germplasm with differences in nitrogen use efficiency. Background Technology

[0002] Sugarcane is an important sugar crop in my country and Guangxi. Excessive application of chemical nitrogen fertilizer has long been a problem in sugarcane production, with application rates reaching 400-600 kg / hm² at times. -2 The nitrogen content is more than three times that of major sugarcane producing regions such as Brazil. However, continuous application of high-nitrogen fertilizers cannot consistently and significantly increase yield and sugar production. Literature reports that the nitrogen recovery rate of sugarcane is mostly less than 15%, and excessive nitrogen fertilizer application can also cause environmental problems such as soil acidification, water eutrophication, and non-point source pollution. Raun and Johnson (1999) demonstrated that a 1% increase in nitrogen use efficiency (NUE) in global grain production can reduce nitrogen fertilizer application by $235 million annually, in addition to environmental and social benefits. Therefore, breeding nitrogen-efficient varieties is an effective way to balance cost, benefit, and environmental factors. The prerequisite for breeding nitrogen-efficient varieties is the availability of practical key physiological or phenotypic indicators. Biomass and nitrogen use efficiency are important physiological indicators for screening nitrogen-efficient sugarcane varieties. However, traditionally, sugarcane biomass and nitrogen use efficiency are generally obtained through destructive sampling. Non-destructive leaf area estimation methods are also used in sugarcane, but they are still mainly measured manually, which is labor-intensive and may cause unintentional damage.

[0003] Automated image analysis can non-destructively, high-throughputly, and with low invasiveness determine plant phenotypic traits. In the past decade, image-based high-throughput phenotyping has gained increasing attention and has been applied to the screening of nitrogen-efficient germplasm in wheat and ryegrass. While the screening systems for wheat and ryegrass provide a reference for screening nitrogen-efficient phenotypes in sugarcane, they cannot be used as the sole basis for sugarcane screening. Due to individual crop morphological differences, such as single-plant morphology, tiller density, leaf overlap, and repeated leaf fall and regrowth, specific phenotypic studies tailored to individual sugarcane conditions are still necessary. Therefore, this invention investigates a method for rapidly screening germplasm with differences in nitrogen use efficiency using phenotypic images. Summary of the Invention

[0004] The purpose of this invention is to provide a rapid phenotypic image method for screening sugarcane germplasm with differences in nitrogen use efficiency. This method, based on the LemnaTech phenotypic platform, obtains two-dimensional phenotypic images. Through image segmentation and extraction, phenotypic trait indicators related to sugarcane are obtained. These phenotypic indicators show a significant positive correlation with sugarcane nitrogen use efficiency and biomass, and can serve as screening indicators for differences in sugarcane nitrogen use efficiency, providing a theoretical reference for sugarcane growth and the breeding of nitrogen-efficient varieties.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A rapid phenotypic image method for screening sugarcane germplasm with differential nitrogen use efficiency includes the following steps:

[0007] S1: Using typical nitrogen-inefficient GT11 and typical nitrogen-efficient GXASF1-01-11 sugarcane germplasms as materials, the germplasms were cultured with nitrogen solutions of different concentrations, and two-dimensional phenotypic images were collected using the LemnaTech 3D high-throughput phenotyping platform at 60d, 120d, and 180d after treatment.

[0008] S2: The acquired images are processed by image processing software to extract phenotypic traits that are related to the sugarcane's response to different nitrogen concentration treatments;

[0009] S3: After collecting the data phenotypic indicators, they were processed using standardized formulas and analyzed after being tested by KMO and Bartlett to screen out key phenotypic indicators for evaluating sugarcane nitrogen use efficiency.

[0010] Standardized formula: zij = (xij - xi) / si

[0011] Where: zij is the standardized variable value; xij is the actual variable value. Xi is the arithmetic mean of each variable (indicator), and si is the standard deviation;

[0012] S4: Based on the selected key phenotypic index data, a regression model between the key phenotypic index and sugarcane biomass and nitrogen use efficiency is established using the multiple linear stepwise regression method. The linear regression equation between sugarcane growth (dry weight and fresh weight) and nitrogen use efficiency and the key phenotypic index is then determined.

[0013] As a further optimization of the present invention: when acquiring images, photos of the top surface, side surface at 0° and 90° of the sample are acquired respectively.

[0014] As a further optimization of the present invention, the image is processed through the following process:

[0015] (1) Using the software LemnaGrid Expert, select the RGB image to be processed, select Demosaicing, and remove edge noise from the image.

[0016] (2) Using LAB to grey convter software, the RGB image was converted to grayscale. The mode that best showed the sugarcane plant in the lightness, green-magenta, and blue-yellow modes was selected and the image was processed.

[0017] (3) Using the Threshold program, select the Advanced mode, then select bright, then select two thresholds in Global threshold, and adjust the threshold to make the plant black against a white background.

[0018] (4) Use the fill area module to adjust the size and remove small areas of the target plant to eliminate background interference.

[0019] (5) Using the Utilities module, select Universal converter and choose the default mode. This process allows for the conversion between object lists, RGB images, and grey images.

[0020] (6) In Image object, under manipulation, select image object composition. This module allows you to treat the different parts of an image as a whole.

[0021] (7) Select the mean color property. This module can calculate the RGB values ​​of the plant and its variance.

[0022] (8) Finally, the image can be output, which is a pure plant image after removing the background. During output, various pixel data can be selected. Using ruler software, pixels can be converted into length and area units.

[0023] As a further optimization of the present invention: the sugarcane-related characterization shape indicators in step S2 include: 0° side surface area, 0° side surface spread, 0° side surface perimeter, 0° side surface compactness, 0° side surface vertical distance, 0° side surface area of ​​the circumscribed polygon, 0° side surface perimeter of the circumscribed polygon, 0° side surface eccentricity, 0° side surface horizontal distance, 0° side surface minimum circumscribed circle diameter, 0° side surface minimum circumscribed circle area, 0° side surface roundness, 0° side surface blue, 0° side surface green, 0° side surface red, 0° side surface hue, 90° side surface area, 90° side surface spread, 90° side surface perimeter, 90° side surface compactness, and 90° side surface vertical distance. Area of ​​the 90° circumscribed polygon, perimeter of the 90° circumscribed polygon, eccentricity of the 90° circumscribed polygon, horizontal distance of the 90° circumscribed polygon, diameter of the minimum circumscribed circle of the 90° circumscribed polygon, area of ​​the minimum circumscribed circle of the 90° circumscribed polygon, roundness of the 90° circumscribed polygon, blue of the 90° circumscribed polygon, green of the 90° circumscribed polygon, red of the 90° circumscribed polygon, hue of the 90° circumscribed polygon, area of ​​the top surface, expansion of the top surface, perimeter of the top surface, compactness of the top surface, vertical distance of the top surface, area of ​​the circumscribed polygon of the top surface, perimeter of the circumscribed polygon of the top surface, eccentricity of the top surface, horizontal distance of the top surface, diameter of the minimum circumscribed circle of the top surface, area of ​​the minimum circumscribed circle of the top surface, roundness of the top surface, blue of the top surface, green of the top surface, red of the top surface, hue of the top surface.

[0024] As a further optimization of the present invention, the key performance indicators are: projected area (side view at 0°, 90° and top view), unfolding degree, vertical distance, minimum circumscribed circle diameter, color (blue, green, red), eccentricity of the top surface, compactness and roundness.

[0025] As a further optimization of the present invention: the linear regression equations are respectively: sugarcane dry weight Y DW =0.618*X1 + 0.261*X2 - 0.145*X3; Fresh weight Y FW =0.586*X1+0.271*X2-0.148*X3; Nitrogen utilization rate YNUE=0.31*X4.

[0026] Note: X1 is the projected area at 90° of the side, X2 is the perimeter at 0° of the side, X3 is the horizontal distance at 0° of the side, and X4 is the area at 0° of the side.

[0027] As a further optimization of the present invention: step S1 uses NH4NO3 as a nitrogen source and cultured in nine different nitrogen solutions of 0.1mM, 0.2mM, 0.4mM, 4mM, 8mM, 12mM, 16mM, 20mM and 24mM.

[0028] As a further optimization of the present invention: the sugarcane germplasm material in step S1 is prepared by first germinating single-bud sugarcane stems at 37-38℃, then moving them to a greenhouse until they grow to one true leaf, and then transplanting them into PVC pots with a diameter of 22 cm and a height of 14 cm, planting one plant per pot. The substrate is mixed in advance, filled into the pots, and thoroughly watered. The amount of substrate used is 80% of the volume of the pot. After planting, the sugarcane is watered with 100 ml of clean water every day. After three true leaves have grown, the sugarcane is alternately watered with 100 ml of nitrogen-containing nutrient solution and 100 ml of clean water every day.

[0029] As a further optimization of the present invention: the matrix is ​​composed of washed sand and vermiculite in a 3:1 volume ratio.

[0030] Compared with the prior art, the advantages and beneficial effects of the present invention include:

[0031] This method, based on the LemnaTech phenotyping platform, obtains high-throughput two-dimensional images of sugarcane phenotypes. After image segmentation and extraction, phenotypic traits related to sugarcane growth and nitrogen use efficiency can be obtained. These phenotypic traits show a significant positive correlation with sugarcane nitrogen use efficiency and biomass, and can serve as screening indicators for differences in sugarcane nitrogen efficiency, providing a theoretical reference for the breeding of sugarcane varieties with high growth and nitrogen use efficiency. The advantages include non-destructive sampling that does not damage the integrity of the sugarcane plant, and the fact that phenotyping does not affect subsequent sugarcane growth. The LemnaTech phenotyping platform enables continuous and automated high-throughput image acquisition, saving labor. Furthermore, the images are acquired through a fixed acquisition chamber on the phenotyping platform, ensuring the stability and consistency of the optical parameters of the acquired images, and guaranteeing the reliability and accuracy of subsequent image processing and phenotypic data. Attached Figure Description

[0032] Figure 1 Original sugarcane image;

[0033] Figure 2 The sugarcane image obtained by image segmentation and extraction;

[0034] Figure 3 To extract various metrics from images. Detailed Implementation

[0035] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0036] I. Acquisition of Sugarcane Germplasm Materials

[0037] The experiment was conducted from January to December 2020 in the glass greenhouse of the Sugarcane Research Institute of the Guangxi Zhuang Academy of Agricultural Sciences. During summer and autumn, the daytime temperature in the greenhouse was 25-35℃, and the nighttime temperature was 20-25℃. During winter, the daytime temperature was 20-35℃, and the nighttime temperature was 15-25℃. The experimental materials were the nitrogen-efficient material GXASF1-08-1-11 and the nitrogen-inefficient material GT11, which were obtained through previous screening. The nitrogen utilization rates of the two materials differed by a factor of 1.5.

[0038] The sugarcane seed stalks were single-bud stalks. The disinfection and germination treatment of the seed stalks were the same as those in the experiment by Yang Liu et al. (2020). The seed stalks were first germinated at 37-38℃, and then moved to a greenhouse. When they grew to one true leaf, they were transplanted into PVC pots with a diameter of 22 cm and a height of 14 cm, with one plant per pot. The substrate was a mixture of washed sand and vermiculite (3:1 volume ratio). The substrate was mixed in advance, filled into the pots, and thoroughly watered. The amount of substrate used was 80% of the volume of the pot. After planting, the sugarcane was watered with 100 ml of clean water every day. After three true leaves had grown, the sugarcane was alternately watered with 100 ml of nitrogen-containing nutrient solution and 100 ml of clean water every day. NH4NO3 was used as the nitrogen source, and nine nitrogen concentration treatments were set up: N0.1, N0.2, N0.4, N4, N8, N12, N16, N20, and N24, representing the molar concentrations of pure nitrogen in the nutrient solution of 0.1, 0.2, 0.4, 4, 8, 12, 16, 20, and 24 mmol / L, respectively. The basic nutrient solution formula was the same as that used in the experiment by Yang Liu et al. (2020). Forty-eight pots were planted in each treatment.

[0039] II. Image Extraction

[0040] Samples were collected after one and two months of growth. At this time, visible growth was observed in the samples. Samples treated with low nitrogen showed stunted growth, small and yellowing leaves, and no stem elongation. Sugarcane treated with high nitrogen showed normal growth, dark green leaves, and normal stem emergence and elongation.

[0041] Image acquisition process: Top, side, and 0° and 90° photos were acquired using the lemnTec phenotyping platform. The RGB cameras used were GT6600c-02-2631d-07193CDEV and GT6600c-02-2631d-07194(DEV-000F31034C49).

[0042] RGB images are processed through the following steps:

[0043] 1. Using the platform's built-in software, lemnaGrid expert, select the RGB image to be processed, choose Demosaicing, and remove edge noise from the image.

[0044] 2. Using LAB to grey convter software, convert the RGB image to grayscale. Select the mode that best highlights the sugarcane plant among the three modes: lightness, green-magenta, and blue-yellow, and process the image accordingly.

[0045] 3. Using the Threshold program, select Advanced mode, then select bright, and in Global threshold, select two thresholds. Adjust the thresholds to make the plant appear as a black background against a white background.

[0046] 4. Using the fill area module, adjust the size to remove small areas of the target plant to eliminate background interference.

[0047] 5. Using the Utilities module, select Universal converter and choose the default mode. This process allows for the conversion between object lists, RGB images, and grey images.

[0048] 6. In Image object manipulation, select image object composition. This module allows you to treat the different parts of an image as a whole.

[0049] 7. Select the mean color property. This module can calculate the RGB values ​​of the plant and its variance.

[0050] 8. Finally, the image can be output, which is a pure plant image after removing the background. During output, various pixel data can be selected. Using ruler software, pixels can be converted into length and area units.

[0051] III. Key Phenotypic Indicators for Determining Nitrogen Use Efficiency in Sugarcane Germplasm

[0052] (1) Phenotypic indicators related to sugarcane growth and nitrogen use efficiency were extracted after image processing, as shown in Table 1.

[0053] Table 1: Sugarcane phenotypic traits collected by the LemnaTec 3D high-throughput phenotyping platform

[0054]

[0055]

[0056] (2) After the data indicators are collected, they are standardized by z-score.

[0057] z-core normalization formula: zij = (xij - xi) / si

[0058] Where: zij is the standardized variable value; xij is the actual variable value. Xi is the arithmetic mean of each variable (indicator), and si is the standard deviation.

[0059] (3) The results of the KMO test and Bartlett's test are as follows:

[0060] The KMO value is 0.875, with a test result >0.5, and the Bartlett's value is 0.000, with a test result <0.05, indicating that the data can be used for principal component extraction.

[0061] (4) Extract factors with initial eigenvalues ​​> 1

[0062] Factors with initial eigenvalues ​​greater than 1 were extracted. A total of 8 common factors were identified from 45 phenotypic traits, with a cumulative variance of 91.18%, indicating that these 8 common factors can explain 91.18% of the total data characteristics.

[0063] The variance contribution rate of common factor 1 is 54.40%. Common factor 1 has a relatively large load on the indicators of projected area (lateral 0, 90 and top surface), spread, vertical distance, and minimum circumscribed circle diameter. These indicators are mainly related to the canopy morphology of sugarcane leaves and can be called sugarcane canopy morphology factors. The variance contribution rates of factors 2 and 3 are 11.40% and 10.75%, respectively. The indicators explained by these two factors are mainly color (blue, green, red), and can be called sugarcane color factors. The cumulative variance contribution rate of factors 4 to 8 is 16.62%. The main indicators they point to are the eccentricity, compactness, and roundness of the top surface. They are mainly related to the plant shape of sugarcane and can be collectively called sugarcane plant shape factors.

[0064] Table 2: Factor analysis results of 15 phenotypic indicators

[0065] Table 4Factor analysis of 15phenotypic characteristics

[0066]

[0067]

[0068] IV. Regression Model Analysis of Phenotypic Indicators with Sugarcane Biomass and Nitrogen Use Efficiency

[0069] Biomass and nitrogen use efficiency are important indicators for evaluating sugarcane growth and nitrogen use efficiency.

[0070] To better explore the relationship between phenotypic indicators such as biomass (dry weight and fresh weight), nitrogen use efficiency, and biomass (dry weight and fresh weight), a multiple linear regression analysis method was introduced to establish a predictive model. After multiple modeling analyses, it was found that the model using data from the 0° and 90° sides had the highest R² value. Therefore, the regression analysis was ultimately conducted using the phenotypic parameters from the 0° and 90° sides. Table 3 shows that three phenotypic indicators from fresh weight and dry weight were included in the model: the projected area A(90°) of the 90° side, the perimeter CL(0°) of the 0° side, and the horizontal distance OEX(0°) of the 0° side. For nitrogen use efficiency, the projected area A(0°) of the 0° side was also included in the model. YFW, YDW, and YNUE represent sugarcane fresh weight (FW), dry weight (DW), and nitrogen utilization rate (NUE), respectively. X1, X2, X3, and X4 represent the projected area A (90°) of the side 90, the perimeter CL (0°) of the side 0, the horizontal distance OEX (0°) of the side 0, and the projected area A (0°) of the side 0, respectively.

[0071] As shown in Table 4, there are three models for both dry weight and fresh weight, with the third model having the largest R-values, at 0.7 and 0.68 respectively. There is only one model for nitrogen utilization rate, with an R-value of 0.31. The Durbin-Watson (DW) test values ​​are between 0.45 and 0.52, close to 0, indicating that the residuals have autocorrelation and the model is usable.

[0072] Table 3: Models for sugarcane dry weight, fresh weight, and single-unit utilization rate

[0073]

[0074] Table 4: Model parameters for sugarcane dry weight, fresh weight, and single-unit utilization rate

[0075]

[0076] Based on the multiple linear regression coefficients of the trademark, the standardized linear regression equations for each Y value are as follows:

[0077] Y DW = 0.618*X1 + 0.261*X2 - 0.145*X3

[0078] Y FW = 0.586*X1 + 0.271*X2 - 0.148*X3

[0079] Y NUE =0.31*X4.

[0080] Note: X1 is the projected area at 90° of the side, X2 is the perimeter at 0° of the side, X3 is the horizontal distance at 0° of the side, and X4 is the area at 0° of the side.

[0081] In summary, the key phenotypic characteristics of sugarcane growth and nitrogen use efficiency obtained by the method of this invention can be used as screening indicators for differences in sugarcane nitrogen efficiency utilization. The regression model established between sugarcane growth and nitrogen rate and phenotypic indicators can be used as a predictive model for sugarcane growth and nitrogen use efficiency, providing a theoretical reference for the breeding of sugarcane varieties with high growth and nitrogen efficiency.

[0082] The above description, in conjunction with specific / preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various substitutions or modifications can be made to these described embodiments without departing from the inventive concept, and all such substitutions or modifications should be considered within the scope of protection of the present invention.

Claims

1. A rapid phenotypic image method for screening sugarcane germplasm with differential nitrogen use efficiency, characterized in that: Includes the following steps: S1: Using typical nitrogen-inefficient GT11 and typical nitrogen-efficient GXASF1-01-11 sugarcane germplasms as materials, sugarcane plants were cultured with nitrogen solutions of different concentrations, and two-dimensional images of sugarcane plant phenotypes were obtained using the LemnaTech 3D high-throughput phenotyping platform at 60d, 120d, and 180d after treatment. S2: The acquired images are processed by image processing software to extract phenotypic traits that are related to the sugarcane's response to different nitrogen concentration treatments; S3: After collecting the data phenotypic indicators, they were processed by standardized formulas and analyzed after being tested by KMO and Bartlett to screen out key phenotypic indicators for evaluating sugarcane nitrogen use efficiency. Standardized formula: zij = (xij - xi) / si Where: zij is the standardized variable value; xij is the actual variable value; Xi is the arithmetic mean of all variables; and si is the standard deviation. S4: Based on the selected key phenotypic index data, a regression model between the key phenotypic index and sugarcane biomass and nitrogen use efficiency is established using the multiple linear stepwise regression method. The linear regression equation between sugarcane growth and nitrogen use efficiency and key phenotypic index is determined, where sugarcane growth is dry weight and fresh weight. The images were captured by taking photos of the top surface of the sample, and the side views at 0° and 90°. The image is processed through the following steps: (1) Using the software lemnaGrid expert, select the RGB image to be processed, select Demosaicing to remove edge noise from the image; (2) Using LAB to grey convter software, the RGB image was converted to grayscale. The mode that best distinguishes the sugarcane plant among the three modes of lightness, green-magenta, and blue-yellow was selected and the image was processed. (3) Using the Threshold program, select the Advanced mode, select bright, select two thresholds in Global threshold, and adjust the threshold to make the plant black against a white background. (4) Use the fill area module to adjust the size and remove small areas of the target plant to eliminate background interference; (5) Use the Utilities module, select Universal converter, and select the default mode. This process can convert between object list, RGB image, and Grey image. (6) In Image object, under manipulation, select image object composition. This module allows the parts of an image to be treated as a whole. (7) Select the mean color property. This module can calculate the RGB values ​​of the plant and its variance. (8) Finally, the image can be output, which is a pure plant image after removing the background; when outputting, you can select the pixel data of various properties; using the ruler software, you can convert the pixels into length and area units; The sugarcane-related shape indicators mentioned in step S2 include: 0° lateral area, 0° lateral spread, 0° lateral perimeter, 0° lateral compactness, 0° lateral vertical distance, 0° lateral circumscribed polygon area, 0° lateral circumscribed polygon perimeter, 0° lateral eccentricity, 0° lateral horizontal distance, 0° lateral minimum circumscribed circle diameter, 0° lateral minimum circumscribed circle area, 0° lateral roundness, 0° lateral blue, 0° lateral green, 0° lateral red, 0° lateral hue, 90° lateral area, 90° lateral spread, 90° lateral perimeter, 90° lateral compactness, 90° lateral vertical distance, and 90° lateral circumscribed polygon area. Polygon area, perimeter of the circumscribed polygon at 90° lateral angle, eccentricity of the circumscribed polygon at 90° lateral angle, horizontal distance at 90° lateral angle, diameter of the minimum circumscribed circle at 90° lateral angle, area of ​​the minimum circumscribed circle at 90° lateral angle, roundness of the circumscribed polygon at 90° lateral angle, blue color of the circumscribed polygon at 90° lateral angle, green color of the circumscribed polygon at 90° lateral angle, red color of the circumscribed polygon at 90° lateral angle, hue of the circumscribed polygon at 90° lateral angle, area of ​​the top surface, expansion of the top surface, perimeter of the top surface, compactness of the top surface, vertical distance of the top surface, area of ​​the circumscribed polygon at 90° lateral angle, perimeter of the circumscribed polygon at 90° lateral angle, eccentricity of the top surface, horizontal distance of the top surface, diameter of the minimum circumscribed circle at 90° lateral angle, area of ​​the minimum circumscribed circle at 90° lateral angle, roundness of the top surface, blue color of the top surface, green color of the top surface, red color of the top surface; The key performance indicators are projected area, unfolding degree, vertical distance, minimum circumscribed circle diameter, color, top surface eccentricity, density, and roundness. The projected area includes the side projection area at 0°, 90°, and the top projection area; Projected area, spread, vertical distance, and minimum circumscribed circle diameter are related to the canopy morphology of sugarcane leaves and are called sugarcane canopy morphology factors; color is called sugarcane color factor; eccentricity, compactness, and roundness of the top surface are related to the plant shape of sugarcane and are collectively called sugarcane plant shape factors. By using scatter plots and biomass (dry weight and fresh weight), nitrogen use efficiency, and exploring the relationship between phenotypic indicators such as biophysical and nitrogen use efficiency of sugarcane, a multiple linear regression analysis method was introduced to establish a predictive model. After multiple modeling analyses, it was found that the model established by introducing data from 0° and 90° of the side surface had the highest R2 value. Therefore, regression analysis was performed using phenotypic parameters from 0° and 90° of the side surface. The linear regression equations are as follows: Sugarcane dry weight YDW = 0.618 X1+0.261 X2-0.145 X3; Fresh weight YFW = 0.586 X1+0.271 X2-0.148 X3; Nitrogen utilization rate YNUE = 0.31 X4; Note: X1 is the projected area at 90° of the side, X2 is the perimeter at 0° of the side, X3 is the horizontal distance at 0° of the side, and X4 is the area at 0° of the side.

2. The method for rapidly screening phenotypic images of sugarcane germplasm with differential nitrogen use efficiency according to claim 1, characterized in that: Step S1 involves using NH4NO3 as the nitrogen source and culturing the cells in nine different nitrogen solutions of 0.1 mM, 0.2 mM, 0.4 mM, 4 mM, 8 mM, 12 mM, 16 mM, 20 mM, and 24 mM.

3. The method for rapidly screening phenotypic images of sugarcane germplasm with differential nitrogen use efficiency according to claim 1, characterized in that: The sugarcane germplasm material mentioned in step S1 is prepared by first germinating single-bud sugarcane stems at 37-38℃, then moving them to a greenhouse until they grow to one true leaf, and then transplanting them into PVC pots with a diameter of 22cm and a height of 14cm, planting one plant per pot. The substrate is mixed in advance, filled into the pots, and thoroughly watered. The amount of substrate used is 80% of the volume of the pot. After planting, the sugarcane is watered with 100ml of clean water every day. After three true leaves have grown, the sugarcane is watered with alternating nitrogen-containing nutrient solution and 100ml of clean water every day.

4. The method for rapidly screening phenotypic images of sugarcane germplasm with differential nitrogen use efficiency according to claim 3, characterized in that: The matrix consists of washed sand and vermiculite in a 3:1 volume ratio.

Images