CN113607734A - Visual method for lossless estimation of chlorophyll content and distribution of plants - Google Patents
Visual method for lossless estimation of chlorophyll content and distribution of plants Download PDFInfo
- Publication number
- CN113607734A CN113607734A CN202110921481.XA CN202110921481A CN113607734A CN 113607734 A CN113607734 A CN 113607734A CN 202110921481 A CN202110921481 A CN 202110921481A CN 113607734 A CN113607734 A CN 113607734A
- Authority
- CN
- China
- Prior art keywords
- chlorophyll content
- image
- plant
- spad
- plants
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N2021/8466—Investigation of vegetal material, e.g. leaves, plants, fruits
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A40/00—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production
- Y02A40/10—Adaptation technologies in agriculture, forestry, livestock or agroalimentary production in agriculture
Abstract
The invention discloses a visual method for lossless estimation of chlorophyll content and distribution of plants, which comprises the steps of collecting plant images by using a visible light camera, identifying all branches of the plants by using a target detection algorithm, selecting target parts by using a rectangular frame, calculating a rectangular frame with the largest height as a main branch area of the plants, and segmenting the main branch area; extracting layered color factors under different color spaces, and performing inversion modeling on a multi-color factor combination and an SPAD value measured by a chlorophyll content measuring instrument to obtain a color factor combination model with the highest fitting degree; the model is applied to plant canopy leaves to represent the distribution of SPAD values, visual display of chlorophyll content on the whole plant plane is realized, photosynthesis of plants is judged through the chlorophyll content, and changes of the chlorophyll content of the plants which are difficult to identify by naked eyes can be observed. The method does not need to damage plants and has objective and accurate estimation result.
Description
Technical Field
The invention relates to the field of image analysis and processing, in particular to a visual method for lossless estimation of chlorophyll content and distribution of plants.
Background
Chlorophyll is a pigment in higher plants and an important basis in plant physiological study, and the content of chlorophyll indicates the growth and health condition of plants, plays an important role in plant photosynthesis and provides energy for the growth and development of the plants. Nitrogen is a component element of chlorophyll, the chlorophyll content is closely related to the nitrogen content of plants in the plant growth process, and the growth process of the plants is influenced to a certain extent by over-high or over-low nitrogen content. Chlorophyll can indirectly reflect the nitrogen content level of a plant, and by utilizing the rapid visualization of the chlorophyll content, the early diagnosis of nitrogen deficiency and excess nitrogen nutrition can be carried out before the stress symptom visible to naked eyes of the plant appears, so that the growth monitoring and the growth judgment of the plant are realized, and further technical guidance is provided for determining and adjusting cultivation management measures, and therefore, the rapid estimation and the visual display of the chlorophyll content have important significance.
In the process of plant growth, when stress occurs, physiological and biochemical phenotypic parameters of plants are changed, but the plants are influenced by the duration of the stress, the stress degree and the plant resistance, and have no obvious difference or change in appearance, so that the change of the plants caused by the stress is difficult to observe by naked eyes. Chlorophyll is a probe for plant photosynthesis, and the determination of the chlorophyll content of plants plays an important guiding role in understanding whether plants grow normally and judging whether plants are influenced by stress. Along with the requirement of large-scale, high-precision and rapid monitoring of plant growth conditions, a plurality of methods for measuring the chlorophyll content of plants appear.
The traditional methods for measuring chlorophyll mainly comprise a spectrophotometry method, a chlorophyll measuring instrument method and the like.
When the spectrophotometry is used, fresh (or dried) plant leaves are taken, cut into pieces (the midrib is removed), evenly mixed and put into a mortar, 80 percent of acetone, calcium carbonate and quartz sand are added, the mixture is ground into homogenate, 80 percent of acetone is added, the homogenate is transferred into a centrifuge tube for centrifugation, after a leaching solution is taken, a visible spectrophotometer is used for color comparison, and the light absorption values at the wavelength of 663nm, 645nm and 652nm are measured, and 80 percent of acetone is used as a reference. Calculating chlorophyll a and chlorophyll b according to a formula, and calculating to obtain the total content (a + b) of chlorophyll. However, the problems of damage to plant growth, multiple detection processes, long period and the like exist when the content of chlorophyll is measured by using a spectrophotometry method.
The chlorophyll meter is used for determining the relative quantity of the current chlorophyll in the leaves by measuring the difference between the optical concentrations of the leaves at two wavelengths, namely 650nm and 940 nm. When chlorophyll measuring instrument is used, a point on the plant leaf is clamped by a chlorophyll content measuring instrument such as SPAD-502, the SPAD value of the point is measured, and the value is used for indicating the chlorophyll content in the current leaf. When the chlorophyll content is measured by using a chlorophyll measuring instrument, the measuring head of the chlorophyll measuring instrument can be used for measuring the chlorophyll content only by ensuring that the sampling blade completely covers the receiving window and clamping a certain sampling area (2mm x 3mm) and not thick (the maximum thickness is 1.2mm), so that the chlorophyll measuring instrument method has poor universality in application, is difficult to measure the chlorophyll content of some small blades, and is easy to cause the problems of overlarge error and the like caused by improper manual operation.
A chlorophyll fluorescence instrument is an analytical instrument used in the field of biology, utilizes the fluorescence phenomenon of chlorophyll, adopts a sensor with extremely high sensitivity and reaction speed, and is assisted with photoelectric pulse design to capture fluorescence data in each time period in the electron transfer process, so that the chlorophyll content can be measured. But the equipment of the chlorophyll fluorescence instrument is expensive and has no universality in use. In the process of measuring the chlorophyll content by using a chlorophyll fluorescence instrument, the leaves need to be picked from the growing plants, and the leaves need to be subjected to dark treatment for a long time before measurement. Therefore, the chlorophyll fluorescence instrument still has the problems of damage to plant growth, complex detection process, long detection period and the like.
In summary, the conventional method for measuring the chlorophyll content can only perform destructive measurement by cutting leaves and plants in vitro in breaking, picking, cutting and other ways, and cannot perform continuous measurement on the same plant. The growth of the plant is a complex and continuous dynamic process and is jointly regulated and controlled by genes and the environment, so that the chlorophyll content of each growth stage of the whole plant is required to be analyzed so as to analyze the time change pattern of the genetic control formed by the phenotypic characters.
In recent years, digital image technology provides a new direction and means for plant chlorophyll content detection, and a method for collecting plant leaf images by using a visible light camera and then analyzing chlorophyll content through image analysis is also available at present. However, most of the existing work of analyzing the chlorophyll content by collecting images of plant leaves needs to extract the leaves from the growing plants, place a single leaf below a visible light camera lens to collect the images, and analyze and process the images to analyze the chlorophyll content. The method inevitably causes irreversible damage to the growth or the leaves of the plant, cannot estimate whether the chlorophyll content changes due to factors such as separation, time, environment and the like after the leaves are picked, and cannot realize the visual distribution of the chlorophyll content canopy of the whole plant.
Therefore, the nondestructive in-vivo chlorophyll measurement technology with non-destructiveness, objectivity, accuracy, high resolution, automation and high efficiency is increasingly required obviously, the change of the chlorophyll content is monitored by directly reflecting the change of the chlorophyll content by using the whole plant image acquired by the visible light camera, and the research on estimating and analyzing the chlorophyll content through the whole plant image color parameters has important significance.
Disclosure of Invention
The invention aims to solve the technical problems of complex process, small processing scale, strong destructiveness, non-visual result and the like of the traditional method for measuring the chlorophyll content of plants, and provides a nondestructive, objective, accurate, high-resolution, automatic and efficient visual method for estimating the chlorophyll content and distribution of plants in a nondestructive manner.
In order to achieve the technical purpose, the technical scheme adopted by the invention is as follows:
a visual method for lossless estimation of chlorophyll content and distribution of plants comprises the following steps:
(1) shooting plants by using a visible light camera, and acquiring a complete plant image;
(2) extracting an image of the pure plant part in the main branch region from the complete plant image;
(3) layering the images extracted in the step (2);
(4) respectively measuring the SPAD (soil and plant analyzer) values of all the leaves in each layer by using a chlorophyll measuring instrument, and respectively calculating the SPAD average value of all the leaves in each layer;
(5) establishing an optimal regression model of the chlorophyll content by using a color analysis method and combining the SPAD average values of all the leaves in each layer;
(6) and estimating the chlorophyll content by using an optimal regression model of the chlorophyll content and visualizing the chlorophyll content.
As a further improved technical solution of the present invention, the step (2) specifically comprises:
(2.1) identifying a complete plant image by using a target detection algorithm so as to identify all branches of the plant, selecting a target part by using rectangular frame frames, calculating the height of each rectangular frame, taking the rectangular frame with the largest height as a main branch area of the plant, and dividing the main branch area;
and (2.2) extracting an interested region from the pure plant parts in the main branch region by using a target detection algorithm and a threshold value of a G channel, and generating a mask of the pure plant parts in the main branch region by using a maximum connected domain method to remove the influence of the environmental background on the extraction of the color factor.
As a further improved technical solution of the present invention, the step (3) specifically comprises:
dividing the image extracted in the step (2) into an upper layer, a middle layer and a lower layer, and judging the division ratio of the upper layer, the middle layer and the lower layer according to the height of the main branch region of the plant.
As a further improved technical solution of the present invention, the step (4) specifically comprises:
measuring SPAD values of all the blades of the upper layer, the middle layer and the lower layer respectively by using a chlorophyll measuring instrument, calculating SPAD average values of all the blades of the upper layer, calculating SPAD average values of all the blades of the middle layer and calculating SPAD average values of all the blades of the lower layer.
As a further improved technical solution of the present invention, the step (5) specifically comprises:
(5.1) converting the image extracted in the step (2) into color spaces RGB (Red, Green, Blue, Red, Green, Blue), HSV (Hue, Saturation, Lightness) and La, b (Lightness, a-star, b-star, Lightness, component from Green to Red, component from Blue to yellow) respectively, and calculating parameter values of color factors of each pixel point in the upper, middle and lower images respectively, wherein the color factors comprise R, G, B, G G, B, G, B, G, B,
H、S、V、L、a、b；
(5.2) calculating the parameter average value of each color factor of all pixel points in the upper-layer image; calculating the parameter average value of each color factor of all pixel points in the middle-layer image; calculating the parameter average value of each color factor of all pixel points in the lower layer image;
(5.3) randomly combining a plurality of color factors, establishing a plurality of groups of color factor combination models, taking the parameter average value of the color factor in each layer of image and the SPAD average value of all leaves in each layer as training data sets, and respectively training the plurality of groups of color factor combination models to obtain a plurality of groups of trained color factor combination models, namely a plurality of groups of regression models of chlorophyll content;
(5.4) determining the coefficient R by the root mean square error RMSE2As an index, the fitting performance of multiple groups of trained regression models of chlorophyll content is evaluated and the regression model of chlorophyll content with the best fitting performance, i.e. the best regression model of chlorophyll content, is determined.
As a further improved technical solution of the present invention, the optimal regression model of chlorophyll content in step (5) is:
Y＝-8.51*lg(G)+11.68*R-26.48*G+18.30*B+2.81*G/R+3.85*G/B+40；
wherein Y is the chlorophyll content estimated by the best regression model for chlorophyll content.
As a further improved technical scheme of the invention, the step (6) comprises the following steps:
(6.1) acquiring and processing the plant image to be detected according to the methods in the step (1) and the step (2);
(6.2) splitting the processed image into three channels of red, green and blue to obtain R, G, B values of each pixel point, and calculating lg (G) and lg (G) of each non-0 pixel point,
And
standardizing all pixel points, calculating a plurality of standardized color factor parameter values, substituting the color factor parameter values into an optimal regression model of chlorophyll content to obtain a gray scale map representing the fitting value of the SPAD, amplifying the fitting value of the SPAD in a pixel point interval, and converting the fitting value of the SPAD into a pseudo-color image of COLORMAP _ JET chromaticity, thereby realizing the visualization of the chlorophyll content. The whole pseudo-color image comprises a gradual change range from dark to light of different colors such as blue-green-yellow-red. Wherein red represents a region with high chlorophyll content, light green and light yellow represent a region with medium chlorophyll content, and blue represents a region with low chlorophyll content. Finally, the chlorophyll content data is converted into visual graphs or images to be displayed on a screen, and the chlorophyll content can be visualized on the whole plant plane.
The method comprises the steps of collecting plant images by using a visible light camera, researching the distribution of chlorophyll content in the whole plant by using an image processing algorithm, establishing an optimal chlorophyll content estimation model and carrying out visualization to obtain an image which visually displays the distribution of the chlorophyll content on the plant, judging the photosynthesis of the plant through the chlorophyll content, and observing the change of the chlorophyll content of the plant which is difficult to identify by naked eyes. In addition, in the plant growth process, the chlorophyll content is closely related to the nitrogen content of the plant, and the nitrogen content is too high or too low, so that the plant growth process is influenced to a certain extent. Chlorophyll can indirectly display the nitrogen content level of a plant, and by utilizing the rapid visualization of the chlorophyll content, the early diagnosis of nitrogen deficiency and nitrogen excess nutrition can be carried out before the occurrence of macroscopic stress symptoms of the plant, so that the growth monitoring and the growth judgment of the plant are realized, and further technical guidance is provided for determining and adjusting cultivation management measures.
The invention has the beneficial effects that:
the visible light camera has low cost, can be widely applied, has simple image acquisition process, and does not need to perform destructive sampling on plants. The method has the advantages that the computer graphics and the image processing technology are utilized for the collected plant images, the complex and redundant chlorophyll content data are converted into visual graphs or images to be displayed on a screen, the visualization of the chlorophyll content in the whole plant is realized, and the visual analysis method for estimating the chlorophyll content and the chlorophyll distribution in a lossless manner is provided for the growth process of the plant.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Fig. 2 is an original plant image collected by a visible light camera, and the original plant image sequentially comprises a control group, a low-fertilizer group, a high-fertilizer group and an over-fertilizer group from left to right.
Fig. 3 is a grayscale diagram of fig. 2.
FIG. 4 is a diagram showing the identification effect of the morphological structure phenotype parameters of the branches.
Fig. 5 is a grayscale diagram of fig. 4.
FIG. 6 (a) is a schematic diagram of the image layering ratio of a plant with a plant height.
FIG. 6 (b) is a schematic diagram showing the image layering ratio of plants of another plant height.
Fig. 7 (a) is a grayscale diagram of fig. 6 (a).
Fig. 7 (b) is a grayscale diagram of fig. 6 (b).
Fig. 8 is a comparison graph of a plant original graph and a visual effect graph of chlorophyll content distributed on the whole plant plane, which are acquired by a visible light camera, and the comparison graph, the low fertilizer group, the high fertilizer group and the over fertilizer group are sequentially arranged from left to right.
Fig. 9 is a grayscale diagram of fig. 8.
Fig. 10 is a schematic diagram of an embodiment of the visualization program of the present invention.
Fig. 11 is a grayscale diagram of fig. 10.
Detailed Description
A more complete appreciation of the invention and many of the attendant disadvantages thereof will be readily obtained as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings and tables, wherein the accompanying drawings and tables illustrated herein provide a further understanding of the invention and form a part thereof.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Fig. 1 to 11 show an example of the application of the present invention to study the chlorophyll content distribution of a dustpan-willow plant. By utilizing the visual analysis method for lossless estimation of the chlorophyll content and distribution of the plants, the optimal estimation model of the chlorophyll content in the willow plant for experiments can be obtained, and the visual result of the chlorophyll content in the willow plant can be obtained.
In the application example, considering that growth of the leaves and the branches of the salix dustpan is random and a deep learning algorithm needs a large amount of data, the invention avoids images with similar shooting angles and selects 2000 images containing growth conditions of various branches of the salix dustpan as a data set, so that the difference of the data set is increased, and 90% of random samples are used for training and 10% of random samples are used for testing.
Step one, an image acquisition stage:
as shown in fig. 2 and 3, the plants are photographed by a visible light camera, and a complete plant image is acquired.
Step two, a chlorophyll visualization stage:
using a target detection algorithm to extract branch parts of the salix dustpan as interested areas to extract phenotypic parameters of morphological structures of branches, selecting the target parts by using rectangular frames, calculating the rectangular frame with the largest height as a main branch area of the plant, and segmenting the main branch area, as shown in fig. 4 and 5. A mask Of pure plant parts Of the main branch Region is generated by extracting ROI (Region Of Interest) with a threshold value (25, 255) Of a G (Green ) channel and acquiring a maximum connected domain.
In the application of the invention to the dustpan willow example test process, the dustpan willow test sample is divided into an upper layer, a middle layer and a lower layer (layering is to divide the image of the pure plant part in the divided main branch region): comparing the correlation between the SPAD value and the color factor according to different layering ratios of a plurality of previous tests to obtain the best layering ratio in the plant growth stage: when the height of the plant at the seedling stage (the height of the plant at the seedling stage is the height of the main branch area) is not more than 35cm, the ratio of the upper layer, the middle layer and the lower layer is 3:4:3, as shown in (a) in fig. 6 and (a) in fig. 7; when the plant height in the seedling stage exceeds 35cm, the ratio of the upper layer, the middle layer and the lower layer is 2:6:2, as shown in (b) in FIG. 6 and (b) in FIG. 7.
The SPAD value of all the leaves of the upper layer, the middle layer and the lower layer is respectively measured by adopting a SPAD-502 chlorophyll content measuring instrument, each leaf is measured for three times at different positions, the average value is taken as a result, and finally the SPAD value of all the leaves of each layer is averaged to be taken as the SPAD value of the leaf of the layer.
By converting the image (the image refers to the image of the pure plant part in the main branch region) into different color space analysis, the characteristics of different color spaces can be effectively utilized to capture the interested characteristic region, and the characterization factor more related to the target value can be more accurately obtained. 3 color spaces RGB, HSV, La, b and 4 common conversion factors based on RGB space are adopted in the experiment process applied to the dustpan and willow example by using the invention for carrying out correlation analysis on the SPAD value. In the same way, the combination of sometimes different color features is converted by a conversion algorithm and is also related to chlorophyll of the leaf, so that R, G, B three color factors are combined and calculated, and the color components capable of representing the green degree are considered and converted into other 4 common color factors, and as shown in table 1, the names of the color factors are extracted from different color spaces. The individual color factors extracted from each layer were then averaged to form a data set ready for regression analysis with the previously measured SPAD values for each layer of leaf, with some of the multi-color factors and SPAD values shown in table 2.
Table 1:
table 2:
considering that the sizes of different color factors may be different greatly, and a larger value easily dominates a target result, so that a regression algorithm cannot learn other characteristics, a data normalization algorithm is used for carrying out non-dimensionalization on numerical data, and original data is transformed into a range with a mean value of 0 and a standard deviation of 1 through a formula (1-1).
In the formula, X is a color factor parameter value obtained by direct extraction and has no dimensional unit; xmean-color factor parameter average, dimensionless units; xstd-standard deviation of color factor parameters, dimensionless units.
In the process of applying the method to a dustpan-willow example test, 580 groups of multicolor factors and SPAD values are obtained as a data set, wherein 90% of multicolor factors and 10% of SPAD values are used for training and testing, and the best model is obtained by training when the model variance regularization degree alpha is set to be 3.2e-6 and the maximum iteration number is set to be 10000.
Determining the coefficient R by the root mean square error RMSE2Is an evaluation indexThe fitting degree of the different color factor combination models and the SPAD index was evaluated to test the fitting performance of the regression model, and the results are shown in Table 3. RMSE represents the average prediction error compared to the measured values, with lower values and higher accuracy. R2The percentage of the measurement variance, which represents the model fitting effect and is interpreted by the algorithm model, has a value range of [0,1 ]]，R2The larger the model, the better the model fit. The statistic calculation formula is as (1-2), (1-3):
in the formula, yrealThe SPAD value of the leaf clamped by the handheld chlorophyll measuring instrument is dimensionless; y ispredThe SPAD value predicted by the multicolor factor correlation model is dimensionless; m is the number of data, and the unit is a group; y ismeanIs the average of all SPAD predicted values, and has no dimension unit.
Table 3:
the fitting degrees of the color factor linear regression models in different color spaces were compared, and as shown in table 3, the models obtained by directly fitting the color factors extracted from the three channels of the color spaces with the SPAD values in the first three groups were all low in fitting degree, R was not high2Mostly 0.58. The fitting performance of H, S, V color factors is best in the test process of applying the dustpan and willow example by using the invention, the error RMSE is only 2.62, and R is2Up to 0.60, is significantly better than other color spaces, but weaker than the set of the latter set of RGB color spacesAnd (5) performing synthesis and transformation. The model 4 adds two color factors of G/R and G/B into the model based on the RGB space color factors to express the ratio of the green component to the red component and the blue component respectively, can better reflect the green degree of the plant leaves, and compared with the model 5 which adds a single color factor G/(R + B), realizes the improvement of the decision coefficient and the reduction of the error on the basis of the original RGB color space model 1.
Table 4:
table 4 shows the nonlinear polynomial regression models constructed by different color factors and the estimation errors, in order to reduce the complexity of the model, a variance selection method is used to filter low variance variables for a plurality of independent variables in the model, and only the variable with the highest correlation is kept as a quadratic term and a logarithmic term. It can be seen that the regression model 14 constructed by the color factors lg (G), R, G, B, G/R, G/B and the regression model 15 constructed by the color factors lg (G), R, G, B, G/(R + B) determine the coefficient R2Both up to 0.73, but the model 14 incorporating the G/R and G/B color factors had less error (RMSE 2.16 ═ R<2.21), so regression model 14: y ═ 8.51 × lg (G) +11.68 × R-26.48 × G +18.30 × B +2.81 × G/R +3.85 × G/B +40 fitted best to the SPAD values, with the most significant regression, as the model fitted for optimal chlorophyll content. The G/R and G/B have been proved to better reflect the green degree of the plant leaves than the G/(R + B) color factor in the linear regression model, and the effects of the quadratic term regression model and the logarithmic term regression model also prove that the G/R and the G/B can be separately adjusted as two parameters, so that the relative green degree of the plant leaves can be better fitted.
By visualizing the chlorophyll content of the image of the salix dustpan sample, the change of the chlorophyll content of the salix dustpan caused by the nitrogen element can be more visual and intuitive, and the influence of the nitrogen element on the chlorophyll content of the plant can be quickly reflected. In the process of applying the method to a dustpan willow example test, an image is divided into three channels of red, green and blue to obtain R, G, B values of each pixel point, lg (G), G/B and G/R of each non-0 pixel point are calculated, all the pixel points are standardized to obtain standardized values of color factors lg (G) and R, G, B, G/R, G/B, and the standardized values are substituted into a regression model 14 which is fitted in the upper section and has the highest correlation: y-8.51 × lg (G) +11.68 × R-26.48 × G +18.30 × B +2.81 × G/R +3.85 × G/B +40, a gray scale representing the fit value of SPAD is obtained, the fit value of SPAD is in the [30,50] interval, and the SPAD interval is enlarged to [0,255] and converted to a pseudo color image of COLORMAP _ JET chromaticity. The whole pseudo-color image comprises a gradual change range from dark to light of different colors such as blue-green-yellow-red. Wherein red represents the region with high chlorophyll content, light green and light yellow represent the region with medium chlorophyll content, and blue represents the region with low chlorophyll content, and the effect is shown in fig. 5. By carrying out visual operation on the whole plant, the method is not limited to numbers any more, and the data related to the chlorophyll content is clearly and visually presented.
Step three, an analysis stage:
as can be seen from the visual effect graphs of fig. 8 and 9, the chlorophyll content is distributed in the whole plant plane, for a single plant of salix dustpan, the SPAD value of the upper layer is usually the lowest, and usually does not exceed 40, because the upper layer of leaf is the youngest and tender, the internal structure is incomplete, the breathing is vigorous, and the chlorophyll content is low; the SPAD value gradually increases from the middle layer to the lower layer, because the chlorophyll content is continuously increased along with the growth of the leaves, the leaf age of the lower leaves is larger, the interior of the structural tissue begins to be damaged, and the photosynthetic rate is lower. The application of the invention to the experimental process of the dustpan-willow example shows that the fluctuation of the SPAD value data table is not obvious, and the result is not visual. The SPAD value obtained by automatically calculating the plant image has the same trend with the change trend of the SPAD value measured by a handheld chlorophyll meter, but the expression is more visual and intuitive.
As can be seen from the visual effect graphs of fig. 8 and 9 showing the distribution of chlorophyll content in the whole plant plane, the chlorophyll content of the dustpan willow plant of example a is overall low, the difference between the chlorophyll content of the upper layer and the chlorophyll content of the lower layer is obvious, and the chlorophyll content of the lower layer is high but in a small range. This indirectly reflects that the willow plant of example a has less nitrogen applied or absorbed, so that nitrogen fertilizer can be applied appropriately during the subsequent growth process. The chlorophyll content of the dustpan-willow plant of example d is overall higher, and the chlorophyll content of the upper layer, the middle layer and the lower layer is not obviously different. By observing the plant image of the salix dustpan collected by the visible light camera, most leaves are in a drooping and wilting state, which indirectly reflects that the salix dustpan plant of the example d has higher nitrogen content or absorbs higher nitrogen content, even has a burning symptom, so nitrogen fertilizer is properly reduced in the subsequent growth process.
By comparing the original plant diagram collected by the visible light camera in fig. 8 or fig. 9 with the visual effect diagram of the distribution of the chlorophyll content in the whole plant plane, the distribution of the chlorophyll content and the changes generated in the growth process of the salix dustpan becomes more visual and intuitive. The chlorophyll visualization method can also be used for rapidly and sensitively qualitatively and quantitatively detecting the change before the change is seen by naked eyes, carrying out quantitative analysis, and playing an important role in monitoring the growth of the plants and evaluating the growth vigor of the plants. In general, the stress symptoms of nitrogen deficiency and nitrogen excess are very delayed, and the original image obtained by visual observation in fig. 8 shows that the plants have no significant difference in applying different nitrogen fertilizers except for the over-fertilized group, but a significant difference is found in the visualization effect graph of chlorophyll content distribution, the fertilization group is larger than the red area (i.e., the area with high chlorophyll content) of the control group, and the red area is enlarged as the nitrogen fertilizer is increased. The quick visualization of the chlorophyll content can be used for carrying out early diagnosis of nutrition such as nitrogen deficiency and nitrogen excess before the visible stress symptom of the plants appears, so that the growth monitoring and the growth judgment of the plants are realized, and further technical guidance is provided for determining and adjusting cultivation management measures.
The above is only one embodiment of the present invention, and is not intended to limit the scope of the present invention. That is, it will be apparent to those skilled in the art that any equivalent changes and modifications can be made without substantially departing from the spirit and scope of the present invention as defined in the appended claims. Therefore, such modifications are also all included in the scope of protection of the present invention.
The present invention provides a method for visually analyzing chlorophyll content and distribution in plants without damage, which is described above with reference to the accompanying drawings. It will be appreciated by persons skilled in the art that the above-described embodiments are merely examples, given for purposes of illustration, and are not intended to be limiting, and that any modifications, equivalents, etc. that fall within the teachings of this application and the scope of the claims should be construed to be encompassed therein.
Claims (7)
1. A visual method for lossless estimation of chlorophyll content and distribution of plants is characterized in that: the method comprises the following steps:
(1) shooting plants by using a visible light camera, and acquiring a complete plant image;
(2) extracting an image of the pure plant part in the main branch region from the complete plant image;
(3) layering the images extracted in the step (2);
(4) respectively measuring the SPAD values of all the leaves in each layer by using a chlorophyll measuring instrument, and respectively calculating the SPAD average value of all the leaves in each layer;
(5) establishing an optimal regression model of the chlorophyll content by using a color analysis method and combining the SPAD average values of all the leaves in each layer;
(6) and estimating the chlorophyll content by using an optimal regression model of the chlorophyll content and visualizing the chlorophyll content.
2. The visual method for non-destructive estimation of chlorophyll content and distribution in plants according to claim 1, wherein: the step (2) specifically comprises the following steps:
(2.1) identifying a complete plant image by using a target detection algorithm so as to identify all branches of the plant, selecting a target part by using rectangular frame frames, calculating the height of each rectangular frame, taking the rectangular frame with the largest height as a main branch area of the plant, and dividing the main branch area;
and (2.2) extracting the interested region from the pure plant parts of the main branch region by using a target detection algorithm and a threshold value of a G channel, and generating a mask of the pure plant parts of the main branch region by using a maximum connected domain method.
3. The visual method for non-destructive estimation of chlorophyll content and distribution in plants according to claim 2, wherein: the step (3) specifically comprises the following steps:
dividing the image extracted in the step (2) into an upper layer, a middle layer and a lower layer, and judging the division ratio of the upper layer, the middle layer and the lower layer according to the height of the main branch region of the plant.
4. The visual method for non-destructive estimation of chlorophyll content and distribution in plants according to claim 3, wherein: the step (4) specifically comprises the following steps:
measuring SPAD values of all the blades of the upper layer, the middle layer and the lower layer respectively by using a chlorophyll measuring instrument, calculating SPAD average values of all the blades of the upper layer, calculating SPAD average values of all the blades of the middle layer and calculating SPAD average values of all the blades of the lower layer.
5. The visual method for non-destructive estimation of chlorophyll content and distribution in plants according to claim 4, wherein: the step (5) specifically comprises the following steps:
(5.1) respectively converting the image extracted in the step (2) into color spaces RGB, HSV and La _ b, and respectively calculating the parameter values of the color factors of each pixel point in the upper layer image, the middle layer image and the lower layer image, wherein the color factors comprise R, G, B, G G,
H、S、V、L、a、b；
(5.2) calculating the parameter average value of each color factor of all pixel points in the upper-layer image; calculating the parameter average value of each color factor of all pixel points in the middle-layer image; calculating the parameter average value of each color factor of all pixel points in the lower layer image;
(5.3) randomly combining a plurality of color factors, establishing a plurality of groups of color factor combination models, taking the parameter average value of the color factor in each layer of image and the SPAD average value of all leaves in each layer as training data sets, and respectively training the plurality of groups of color factor combination models to obtain a plurality of groups of trained color factor combination models, namely a plurality of groups of regression models of chlorophyll content;
(5.4) determining the coefficient R by the root mean square error RMSE2As an index, the fitting performance of multiple groups of trained regression models of chlorophyll content is evaluated and the regression model of chlorophyll content with the best fitting performance, i.e. the best regression model of chlorophyll content, is determined.
6. The visual method for non-destructive estimation of chlorophyll content and distribution in plants according to claim 5, wherein: the optimal regression model of the chlorophyll content in the step (5) is as follows:
Y＝-8.51*lg(G)+11.68*R-26.48*G+18.30*B+2.81*G/R+3.85*G/B+40；
wherein Y is the chlorophyll content estimated by the best regression model for chlorophyll content.
7. The visual method for non-destructive estimation of chlorophyll content and distribution in plants according to claim 6, wherein: the step (6) comprises the following steps:
(6.1) acquiring and processing the plant image to be detected according to the methods in the step (1) and the step (2);
(6.2) splitting the processed image into three channels of red, green and blue to obtain R, G, B values of each pixel point, and calculating lg (G) and lg (G) of each non-0 pixel point,
And
standardizing all pixel pointsAnd calculating a plurality of standardized color factor parameter values, substituting the color factor parameter values into the optimal regression model of the chlorophyll content to obtain a gray scale map representing the fitting value of the SPAD, amplifying the fitting value of the SPAD in a pixel point interval, and converting the fitting value of the SPAD into a pseudo-color image of COLORMAP _ JET chromaticity, thereby realizing the visualization of the chlorophyll content.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110921481.XA CN113607734B (en) | 2021-08-11 | 2021-08-11 | Visualization method for nondestructively estimating chlorophyll content and chlorophyll distribution of plant |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110921481.XA CN113607734B (en) | 2021-08-11 | 2021-08-11 | Visualization method for nondestructively estimating chlorophyll content and chlorophyll distribution of plant |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113607734A true CN113607734A (en) | 2021-11-05 |
| CN113607734B CN113607734B (en) | 2023-05-26 |
Family
ID=78308288
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110921481.XA Active CN113607734B (en) | 2021-08-11 | 2021-08-11 | Visualization method for nondestructively estimating chlorophyll content and chlorophyll distribution of plant |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113607734B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080239293A1 (en) * | 2005-11-21 | 2008-10-02 | Fuchigami Leslie H | Portable Meter to Measure Chlorophyll, Nitrogen and Water and Methods |
| US20110125477A1 (en) * | 2009-05-14 | 2011-05-26 | Lightner Jonathan E | Inverse Modeling for Characteristic Prediction from Multi-Spectral and Hyper-Spectral Remote Sensed Datasets |
| WO2014100856A1 (en) * | 2012-12-24 | 2014-07-03 | University Of Technology, Sydney | An image processing based method to estimate crop requirements for nutrient fertiliser |
| CN108982377A (en) * | 2018-07-27 | 2018-12-11 | 内蒙古大学 | Corn growth stage spectrum picture and chlorophyll content correlation and period division methods |
| CN109827957A (en) * | 2018-12-17 | 2019-05-31 | 南京邮电大学 | A kind of rice leaf SPAD value estimating and measuring method based on computer vision and system |
-
2021
- 2021-08-11 CN CN202110921481.XA patent/CN113607734B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080239293A1 (en) * | 2005-11-21 | 2008-10-02 | Fuchigami Leslie H | Portable Meter to Measure Chlorophyll, Nitrogen and Water and Methods |
| US20110125477A1 (en) * | 2009-05-14 | 2011-05-26 | Lightner Jonathan E | Inverse Modeling for Characteristic Prediction from Multi-Spectral and Hyper-Spectral Remote Sensed Datasets |
| WO2014100856A1 (en) * | 2012-12-24 | 2014-07-03 | University Of Technology, Sydney | An image processing based method to estimate crop requirements for nutrient fertiliser |
| CN108982377A (en) * | 2018-07-27 | 2018-12-11 | 内蒙古大学 | Corn growth stage spectrum picture and chlorophyll content correlation and period division methods |
| CN109827957A (en) * | 2018-12-17 | 2019-05-31 | 南京邮电大学 | A kind of rice leaf SPAD value estimating and measuring method based on computer vision and system |
Non-Patent Citations (13)
| Title |
|---|
| 吴少俊: "基于计算机视觉的水稻叶绿素含量测定", 《农机化研究》 * |
| 宋一帆等: "基于RGB模型的大豆叶片叶绿素含量预测", 《江汉大学学报(自然科学版)》 * |
| 张建等: "消费级近红外相机的水稻叶片叶绿素(SPAD)分布预测", 《光谱学与光谱分析》 * |
| 张建等: "针对水稻植株三维叶绿素(SPAD)信息获取的最优波段选择问题研究", 《光谱学与光谱分析》 * |
| 张慧春等: "植物表型平台与图像分析技术研究进展与展望", 《农业机械学报》 * |
| 李岚涛等: "应用数字图像技术进行水稻氮素营养诊断", 《植物营养与肥料学报》 * |
| 李杨先，张慧春，杨旸: "一种基于图像处理技术的植物形态表型参数获取方法", 《林业工程学报》 * |
| 杜玉红等: "基于计算机视觉技术的玉米叶绿素含量检测研究", 《农机化研究》 * |
| 胡阳等: "基于数码影像的香樟树叶片叶绿素分布可视化研究", 《信阳农林学院学报》 * |
| 贺英等: "基于数码相机的玉米冠层SPAD遥感估算", 《中国农业科学》 * |
| 边黎明等: "表型技术在林木育种和精确林业上的应用", 《林业科学》 * |
| 龙耀威等: "基于作物谱图特征的植株分割与叶绿素分布检测", 《光谱学与光谱分析》 * |
| 龙耀威等: "镀膜型光谱成像数据提取与作物叶绿素分布探测研究", 《光谱学与光谱分析》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113607734B (en) | 2023-05-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | High throughput analysis of leaf chlorophyll content in sorghum using RGB, hyperspectral, and fluorescence imaging and sensor fusion | |
| EP2376897B1 (en) | Method for diagnosis of stresses and diseases in higher plants | |
| CN107796764B (en) | Method for constructing wheat leaf area index estimation model based on three-band vegetation index | |
| Mao et al. | Nondestructive measurement of total nitrogen in lettuce by integrating spectroscopy and computer vision | |
| CN110763698A (en) | Hyperspectral citrus leaf disease identification method based on characteristic wavelength | |
| Luo et al. | Detecting aphid density of winter wheat leaf using hyperspectral measurements | |
| Pinto et al. | Classification of Hass avocado (persea americana mill) in terms of its ripening via hyperspectral images | |
| CN109827957B (en) | Rice leaf SPAD value estimation method and system based on computer vision | |
| CN105021529A (en) | Spectrum information and image information fusing crop plant disease and insect pest identifying and distinguishing method | |
| Patane et al. | Chlorophyll and nitrogen estimation techniques: A Review | |
| CN106841167A (en) | The lossless detection method of garden stuff pesticide residue | |
| Motonaga et al. | A method of making digital fruit color charts for cultivation management and quality control | |
| CN102184389B (en) | Automatic annual ring identification method used for tree species with high chromatic aberration of heartwood and sapwood | |
| CN113607734B (en) | Visualization method for nondestructively estimating chlorophyll content and chlorophyll distribution of plant | |
| Zhou et al. | Hyperspectral imaging technology for detection of moisture content of tomato leaves | |
| Summerer et al. | High throughput plant phenotyping: a new and objective method to detect and analyse the biostimulant properties of different products | |
| Tosin et al. | Precision maturation assessment of grape tissues: Hyperspectral bi-directional reconstruction using tomography-like based on multi-block hierarchical principal component analysis | |
| Hippola et al. | Machine learning based classification of ripening and decay stages of Mango (Mangifera indica L.) cv. Tom EJC | |
| Zhang et al. | Evaluation of lawn color and chlorophyll concentration using hyperspectral index | |
| Ali et al. | Comparative analysis of different methods of leaf area index estimation of Strawberry under Egyptian condition | |
| Chen et al. | Prediction of chlorophyll and anthocyanin contents in purple lettuce based on image processing | |
| CN114022389B (en) | Method for distinguishing plant leaf color grade based on digital camera | |
| Rodriguez et al. | New Features and Many Improvements to Analyze Morphology and Color of Digitalized Plant Organs Are Available in Tomato Analyzer 3.0. | |
| Li et al. | Assessment of phytoplankton class abundance using in vivo synchronous fluorescence spectra | |
| Zhang et al. | Estimation of the Chlorophyll Concentration in Sorghum Using Three High Throughput Phenotyping Imaging Techniques |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |