A phenological monitoring method for woody ornamental plant germplasm resource bank

Abstract

The present invention discloses a phenological monitoring method for a woody ornamental plant germplasm resource bank, which belongs to the technical field of phenological monitoring, and includes the following steps: image acquisition and preprocessing of the germplasm resource bank; boundary segmentation and numbering of germplasm plots in the resource bank; acquisition of time series images: plot extraction and color parameter extraction; phenological period classification and image result storage. The image samples and color parameter data of each phenological period within the growth cycle of each germplasm year are stored in the phenological library, and the abnormal growth images and color parameters are stored in the abnormal library. Then, during the subsequent use of the monitoring system, the color parameters of the vegetation in each plot are extracted, and the phenological period of the plant is determined by cluster analysis based on the color parameters. Phenological changes are monitored by comparing changes in monitoring values. The present invention is based on an imaging system of an unmanned aerial vehicle, and utilizes color parameters to establish a phenological monitoring method for a woody ornamental plant germplasm resource bank, which can quickly and conveniently monitor the phenological period of woody plants.

Description

Technical Field

[0001] The invention belongs to the technical field of phenological monitoring, and in particular relates to a phenological monitoring method for a woody ornamental plant germplasm resource bank. Background Art

[0002] Woody ornamental plant germplasm resources are fundamental and strategic resources for the development of the garden industry. Strengthening the collection, preservation, and rational utilization of woody ornamental plant germplasm resources, reserving abundant breeding materials for woody ornamental plant variety innovation, and achieving the localization of major woody ornamental plant varieties are of great significance to meeting the needs of the modern garden industry. The establishment of a germplasm resource bank is an important means to effectively protect and rationally utilize flower germplasm resources, promote the innovation of woody ornamental plant varieties, and promote the development of the garden industry.

[0003] The collection and preservation of forest germplasm resources is a long-term undertaking, requiring significant human, material, and financial resources at every stage, from resource surveys to collection and preservation. Forest germplasm gene banks are large, contain a wide variety of germplasms, and exhibit significant differences in various germplasm traits. Currently, most resource banks still rely on traditional, in-person surveys for growth monitoring and trait surveys, which is extremely labor-intensive and time-consuming. Furthermore, due to the similar phenological periods of germplasms, traditional methods can easily miss optimal observation periods for certain traits, hindering data collection. Some experienced surveyors may use phenological data recorded in previous years to narrow the scope of a single survey. However, phenology is subject to significant fluctuations due to environmental factors, particularly for ornamental plants. Therefore, finding key phenological points and growth anomalies for various traits in a time-saving and labor-efficient manner is crucial for conducting data surveys at germplasm resource banks and for the gardening of germplasm.

[0004] UAV-based imaging systems have been proven to be able to perform plant phenotyping tasks well. In forestry research, they are mainly used in diversity surveys, cover calculations, and biomass predictions. There are few reports on their application in phenological monitoring of ornamental woody flowers.

[0005] Existing phenological remote sensing monitoring methods often use growth curves or vegetation indices to define phenological periods, making them inappropriate for the woody flower resource collections targeted by this invention. Firstly, resource collections are generally composed of plants within a genus, whose growth curves vary minimally. Secondly, the color variations of flowers, leaves, and fruits in woody flowers can complicate decisions when defining phenological periods based on vegetation indices. Summary of the Invention

[0006] The technical problem solved by the present invention is to provide a phenological monitoring method for a woody ornamental plant germplasm resource bank using color parameters, so as to realize rapid and convenient monitoring of the phenological period of woody plants.

[0007] Technical solution: In order to solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0008] A method for monitoring phenology of a woody ornamental plant germplasm resource bank comprises the following steps:

[0009] Step 1: Germplasm resource bank image acquisition and preprocessing;

[0010] Step 2: Boundary division and numbering of seed plots in the resource library;

[0011] Step 3: Get time series images:

[0012] Step 4: Land extraction and color parameter extraction;

[0013] Step 5: Phenological period classification and image result storage.

[0014] Furthermore, in step 1, an aerial image of the germplasm resource bank is obtained by vertically shooting with a fixed height using a drone, and the image is ortho-mosaiced using the drone's built-in remote sensing image processing system, and then a reflectance image is obtained after radiation correction.

[0015] Furthermore, in step 2, a manual drawing method is used to segment the cell boundaries. The Quantum GIS vector operation open source library is used to add the orthophoto map as the base map, and then a new polygon feature layer is created. Then, each polygon feature is drawn according to the land plot boundary of the resource library on the base map. One land plot is a polygon feature, and one layer contains multiple polygon features.

[0016] Furthermore, in step 3, different fixed time intervals are set for different phenological periods according to the characteristics of the tree species. Flowering tree species should be photographed at intervals of 1 day or half a day during their flowering phenological period, and the time intervals should be set to several days during non-important phenological periods.

[0017] Furthermore, in step 4, plot extraction and color parameter extraction are performed; the CIE Lab color parameters of vegetation in each plot are extracted using the drone's built-in remote sensing image processing system or Photoshop, where L represents lightness; a represents the component from green to red, and b represents the component from blue to yellow.

[0018] Furthermore, in step 5, the images of different phenological periods and growth abnormalities of various qualities are classified, and the image samples and color parameter data of each phenological period in the annual growth cycle of various qualities are stored in the phenological library, and the growth abnormality images and color parameters are stored in the growth abnormality library. Based on the extracted color parameters, the three-parameter spatiotemporal distribution map of the colors of various qualities in the phenological period is established in each library.

[0019] Furthermore, in step 5, △E1 between different sampling points in each plot is first calculated.

[0020] △E1=[(△L) 2 +(△a) 2 +(△b) 2 ] 1 / 2

[0021] △L represents the brightness / black and white difference, △L=L 采样点x1 -L 采样点x2 ;

[0022] △a represents the difference between red and green, △a=a 采样点x1 -a 采样点x2 ;

[0023] △b represents the difference between yellow and blue, △b=b 采样点x1 -b 采样点x2 ;

[0024] Then, calculate the color difference value △E2 during different phenological periods under normal conditions:

[0025] △E2=[(△L) 2 +(△a) 2 +(△b) 2 ] 1 / 2 .

[0026] △L represents the brightness / difference between black and white: △L=L 后一时期物候 -L 前一时期物候 ();

[0027] △a represents the difference between red and green: △a=a 后一时期物候 -a 前一时期物候 ;

[0028] △b represents the difference between yellow and blue: △b=b 后一时期物候 -b 前一时期物候 .

[0029] Furthermore, in the subsequent use of the monitoring system, the vegetation color parameters of each plot are extracted, and the phenological period is determined by cluster analysis based on the color parameters. Phenological changes are monitored by comparing the changes in the color difference values ​​△E1' between different sampling points on the photos of a certain texture plot and the color difference values ​​△E2' between the photos of a certain texture plot and the last photo of the same texture plot taken.

[0030] Furthermore, if △E1'>△E 1max If it indicates that there are inconsistencies in phenological periods, alternate years of flowering and fruiting, or abnormal growth among different individual plants of the germplasm, a prompt message will be generated and sent to the investigator's communication device;

[0031] If △E2' ≈ △E 2min , indicating that the phenological change is in a critical state, and a prompt message is generated and sent to the investigator's communication device;

[0032] △E2' > △E 2min , indicating that the phenological stage spans a large distance, sending a warning message;

[0033] △E1' represents the color difference between different sampling points on a certain texture block photograph taken during the subsequent use of the monitoring system;

[0034] △E2' represents the color difference between the photograph of a certain patch of soil taken during the subsequent use of the monitoring system and the last photograph of the same patch taken.

[0035] △E 1max Indicates the maximum color difference between different sampling points on a certain patch of land in the phenology database

[0036] △E 2min Indicates the minimum color difference between adjacent phenological photos of a certain quality in the phenological database.

[0037] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0038] The phenological monitoring method for woody ornamental plant germplasm resource banks of the present invention can quickly obtain pictures of the current status of germplasm through aerial photography, automatically interpret the phenology of the germplasm in combination with color parameters, and can also reflect the characteristics of phenological changes through the size of color difference parameters, and generate prompts or warning information. This method saves time and effort and efficiently finds various key phenological points and growth anomalies of germplasm, quickly and conveniently monitors phenological periods, and is conducive to the development of germplasm resource bank data surveys. There are currently no research reports on the monitoring of specific flowering stages. Traditional technologies mainly rely on manual inspections. When the number of varieties is large, it is easy to miss data collection at important flowering stages. This method is particularly sensitive and convenient in spatiotemporal positioning of variety flowering stages. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a flow chart of the phenological monitoring method for woody ornamental plant germplasm banks;

[0040] Figure 2 This is an image of the crabapple germplasm gene bank taken at an aerial altitude of 15 meters, obtained in an embodiment of the present invention;

[0041] Figure 3 This is an image of the crabapple germplasm gene bank taken at an aerial altitude of 50 meters obtained in an embodiment of the present invention;

[0042] Figure 4 This is an image of the crabapple germplasm gene bank taken at an aerial altitude of 100 meters, obtained in an embodiment of the present invention;

[0043] Figure 5 This is an image of the crabapple germplasm gene bank after plot extraction and color parameter extraction according to an embodiment of the present invention;

[0044] Figure 6 This is a dynamic aerial photo of the phenology of the "Gaoyuan Red" crabapple (from left to right: budding stage, leaf expansion stage, bud stage, early flowering stage, full flowering stage, late flowering stage, leaf fall stage, fruiting stage, and dormancy stage);

[0045] Figure 7 This is a spatiotemporal distribution map of three color parameters extracted from the phenological dynamic aerial photography of the 'Gaoyuanhong' crabapple.

[0046] Figure 8 These are aerial photos of the 'Gaoyuanhong' crabapple at different times;

[0047] Figure 9 This is the cluster analysis result of “Gaoyuan Red” crabapple based on color parameters. DETAILED DESCRIPTION

[0048] The present invention will be further illustrated below with reference to specific examples. The examples are implemented based on the technical solutions of the present invention. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention.

[0049] Standardized germplasm gene banks are typically planted in a "grid" pattern, which also facilitates plant extraction. Therefore, this invention uses a drone imaging system and color parameters to establish a phenological monitoring method for woody ornamental plant germplasm resource banks, enabling rapid and convenient monitoring of woody plant phenological periods.

[0050] like Figure 1 As shown, the phenological monitoring method of the woody ornamental plant germplasm resource bank of the present application mainly includes the following steps:

[0051] Step 1: Germplasm resource bank image acquisition and preprocessing

[0052] Aerial images of the germplasm resource bank were obtained by using a drone to shoot vertically at a fixed height. The images were ortho-mosaiced using the drone's built-in remote sensing image processing system, and reflectance images were obtained after radiation correction.

[0053] Step 2: Boundary division and numbering of seed plots in the resource library

[0054] Considering the accuracy of subsequent information extraction and the reusability of a single resource library seed plot layer, we used manual drawing to segment plot boundaries. Using the Quantum GIS vector operation open-source library, we added an orthophoto as the basemap and created a new polygon feature layer. Each polygon feature was then drawn according to the resource library seed plot boundaries on the basemap. Each plot is a polygon feature, and a layer can contain multiple polygon features. Each plot represents a single variety (germplasm), and the plots (polygons) are numbered sequentially according to row and column numbers.

[0055] Plots are divided according to variety type, with each variety being divided into one block. There is no fixed rule for numbering; Arabic numerals and other numbers are acceptable, but each plot has a unique number.

[0056] Step 3: Obtain time series images according to steps 1 and 2

[0057] Use drones to capture aerial images of the germplasm resource bank within the growth cycle of various germplasm species at fixed time intervals and a fixed altitude (50m-100m is ideal);

[0058] Depending on the characteristics of the tree species, the fixed time intervals for different phenological periods can be different. For example, flowering tree species should be photographed at shorter time intervals (such as 1 day or half a day) during their flowering phenological period (depending on the rainy and snowy weather conditions). The time interval can be set to several days during non-important phenological periods.

[0059] Step 4: Land extraction and color parameter extraction:

[0060] Use the drone's built-in remote sensing image processing system or Photoshop to extract vegetation color parameters (CIELab) within each plot. L represents lightness, ranging from 0 to 100; a represents the green-to-red component, ranging from -128 to 127; and b represents the blue-to-yellow component, ranging from -128 to 127. Common methods for extracting color parameters from remote sensing images are RGB or HSL. CIELab has a wider color gamut, is device-independent, and is based on physiological characteristics. Its unique advantage lies in emphasizing color variations. The parameters are easily accessible, and the method is easily accessible and widely used.

[0061] Step 5: Phenological period classification and image storage:

[0062] Manual classification of images of various phenological periods and growth anomalies for each quality is performed. Image samples and color parameter (CIE Lab) data for each phenological period within each quality's annual growth cycle are stored in the phenological database. Images and color parameters (CIE Lab) of growth anomalies (e.g., pests and diseases) are stored in the growth anomaly database. Based on the extracted color parameters, spatiotemporal distribution maps of the three color parameters for each quality's phenological period are created in each database for intuitive analysis.

[0063] Calculate △E1 between different sampling points in each plot as follows:

[0064] △E1=[(△L) 2 +(△a) 2 +(△b) 2 ] 1 / 2

[0065] In the formula, △L represents the brightness / black and white difference, △L=L 采样点x1 -L 采样点x2 ;

[0066] △a represents the difference between red and green, △a=a 采样点x1 -a 采样点x2 ;

[0067] △b represents the difference between yellow and blue, △b=b 采样点x1 -b 采样点x2 .

[0068] Calculate the color difference value △E2 during different phenological periods under normal conditions of various qualities. The calculation method is as follows:

[0069] △E2=[(△L) 2 +(△a) 2 +(△b) 2 ] 1 / 2

[0070] Where △L represents the brightness / black and white difference: △L=L 后一时期物候 -L 前一时期物候 ;

[0071] △a represents the difference between red and green: △a=a 后一时期物候 -a 前一时期物候 ;

[0072] △b represents the difference between yellow and blue: △b=b 后一时期物候 -b 前一时期物候 .

[0073] Definition of phenological period: Phenological period includes bud break period, leaf expansion period, flowering period (bud period, initial flowering period, peak flowering period, and final flowering period), fruiting period, leaf fall period, and dormancy period (for deciduous tree species).

[0074] During the subsequent use of the monitoring system, the vegetation color parameters (CIE Lab) of each plot are extracted, the phenological period is determined by cluster analysis based on the color parameters, and the phenological changes are monitored by comparing the changes in the △E1' and △E2' values. △E1' represents the color difference between different sampling points on a certain texture plot photographed during the subsequent use of the monitoring system. △E2' represents the color difference between a certain texture plot photographed during the subsequent use of the monitoring system and the last photograph of the same texture plot taken. △E 1max Indicates the maximum color difference between different sampling points on a certain patch of land in the phenology database; △E 2min This value represents the minimum color difference between adjacent phenological photos of a given species in the phenology database. The plant status of △E1' and △E2' is unknown and under test. △E1' and △E2' are calculated using the formula for △E1 and △E2 above. △E1 and △E2 are the color difference values ​​of the phenological period under normal conditions.

[0075] When the present invention is used for the first time, that is, according to steps 1-5, a phenological database is established. It is necessary to manually establish the phenological database based on normal phenological photos and determine △E 1max、 △E 2min When used for subsequent monitoring, the images obtained from daily aerial monitoring are processed, including image ortho-mosaicing and radiation correction, plot boundary segmentation and numbering, plot extraction and color parameter extraction. The methods refer to the methods in steps 1-4. Then the obtained parameters are compared with the data of the monitoring system. No manual judgment is required. Only the size of △E1' and △E2' and △E 1max、 △E 2min Use the system to perform automatic comparisons.

[0076] If △E1'>△E 1max， This indicates that there are inconsistencies in phenological periods or alternate years of flowers and fruits or abnormal growth among different individual plants of this germplasm, and a prompt message is generated and sent to the investigator's communication device.

[0077] △E2'≈△E 2min , indicating that the phenological change is in a critical state, and a prompt message is generated and sent to the investigator's communication device;

[0078] △E2'>△E 2min , indicating that the phenological stage spans a large distance, sending a warning message.

[0079] △E1 and △E2 vary between varieties (depending on their color), so a database of △E1 and △E2 values ​​for each variety should be created when first using this method. On subsequent uses, △E1' and △E2' values ​​should be compared with the database's △E1 and △E2 values. Values ​​outside the normal range indicate significant phenological changes. When values ​​exceed the normal range, manual identification can be performed to determine actual abnormalities. This method is primarily used to monitor phenology and issue alerts if any abnormalities are detected.

[0080] This application establishes a phenological database by counting and collecting phenological data and images of the monitored objects within one year. When used subsequently, the real-time images will be processed and then compared with the data of the monitoring system to automatically determine the phenological changes and generate prompt information to be sent to the investigator's communication equipment. At the same time, the abnormal data determined during the monitoring process will be stored in the abnormal database. The system quickly obtains pictures of the current status of germplasm through aerial photography, and automatically interprets the phenology of the germplasm in combination with color parameters. In addition, it can also reflect the characteristics of phenological changes through the size of color difference parameters and generate prompts or warning information. This method saves time and effort and efficiently finds various key phenological points and growth anomalies of germplasm, quickly and conveniently monitors phenological periods, and is conducive to the development of data surveys of germplasm resource banks.

[0081] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method described in the above embodiment 1 are executed.

[0082] This example takes the Begonia germplasm gene bank as an example:

[0083] Crabapple is a deciduous small tree. In southern Jiangsu, China, the budding period of most varieties is generally early March, the leaf expansion period is mid-to-late March, the flowering period is generally mid-to-late March to late April (each variety takes about 2 weeks from the initial flowering period to the final flowering period), the fruiting period is from May to January of the following year, the leaf fall period is from July to the end of October (there are large differences among different varieties), and the dormancy period is from November to mid-to-late February.

[0084] Taking the 'Gaoyuanhong' crabapple as an example, plot number S1, as shown in 2-4, is the crabapple germplasm gene bank image obtained in this embodiment, specifically the aerial photo of the Yangzhou International Crabapple Germplasm Gene Bank, as shown in Figure 4 Shown is an aerial photo of the phenological dynamics of the 'Gaoyuanhong' crabapple (from left to right: bud break, leaf expansion, bud stage, initial flowering stage, peak flowering stage, final flowering stage, leaf fall, fruiting stage, and dormancy stage).

[0085] First, the image acquisition and image preprocessing of the crabapple germplasm gene bank were carried out, and then the plot extraction and color parameter extraction were carried out, such as Figure 5 The result of plot extraction.

[0086] According to the algorithm in this application, the ΔE1 between different sampling points of the 'Gaoyuanhong' crabapple in plot S1 during each phenological period was calculated. ΔE1 during budding period was 3.3-9.8, ΔE1 during leaf expansion period was 4.5-14.6, ΔE1 during budding period was 10-24.4, ΔE1 during early flowering period was 3.1-15.5, ΔE1 during full flowering period was 13.0-43.2, ΔE1 during late flowering period was 6-17.5, ΔE1 during leaf fall period was 3.3-16.4, ΔE1 during fruiting period was 5.9-14.2, and ΔE1 during dormancy period was 4-16.5.

[0087] Then the color differences of 'Gaoyuanhong' crabapple during each phenological period were calculated: △E2 between budding period and leaf expansion period=40.6~55.2, △E2 between leaf expansion period and bud period=21.7~42.6, △E2 between bud period and initial flowering period=9.6~18.5, △E2 between initial flowering period and full flowering period=8.8~26.8, △E2 between full flowering period and final flowering period=33.3~48.4, △E2 between final flowering period and leaf fall period=14.7~26.9, △E2 between leaf fall period and fruiting period=13.7~28.3, △E2 between fruiting period and dormancy period=28.6~48.1, △E2 between dormancy period and budding period=62.3~78.8.

[0088] Verification: Taking the ‘Gaoyuanhong’ crabapple as an example, we randomly selected two aerial photos taken on different dates, such as Figure 8 As shown. Extract color parameters: A Figure L mean =28.25, a mean =15.5, b mean =-3; Figure B L=48.5,a mean =23.5, b mean =-3.75. Cluster analysis found that the image on March 22nd was clustered with the bud stage, and the image on April 10th was clustered with the full bloom stage. The results are consistent with the actual situation. The clustering results are as follows Figure 9 As shown. Between Figures A and B, △E2' =21.8>△E 2蕾-初开 =9.6~18.5, indicating that the phenology of 'Gaoyuanhong' crabapple changed greatly between the two aerial photography times, and the conclusion is consistent with the actual situation.

[0089] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A method for monitoring phenology of a woody ornamental plant germplasm resource bank, characterized in that: The following steps are involved: Step 1: Germplasm resource bank image acquisition and preprocessing; Step 2: Boundary division and numbering of seed plots in the resource library; Step 3: Get time series images: Step 4: Plot extraction and color parameter extraction: Use the drone's built-in remote sensing image processing system or Photoshop to extract the vegetation color parameters CIE Lab within each plot, where L represents lightness; a represents the component from green to red, and b represents the component from blue to yellow. Step 5: Phenological period classification and image result storage: Classify the different phenological periods and growth abnormality images of various qualities, store the image samples and color parameter data of each phenological period in the growth cycle of various qualities in the phenological database, and store the growth abnormality images and color parameters in the growth abnormality database. Based on the extracted color parameters, establish the three-parameter spatiotemporal distribution map of the color of various qualities in each database; In step 5, the color difference value △E1 between different sampling points in each plot is first calculated. △E1=[(△L1) 2 +(△a1) 2 +(△b1) 2 ] 1 / 2 ; △L1 represents the brightness / black and white difference between different sampling points in the same plot, △L1=L 采样点x1 -L 采样点x2 ; △a1 represents the red / green difference between different sampling points in the same plot, △a1=a 采样点x1 -a 采样点x2 ; △b1 represents the yellow / blue difference between different sampling points in the same plot, △b1=b 采样点x1 -b 采样点x2 ; Then calculate the color difference value △E2 during different phenological periods under normal conditions: △E2=[(△L2) 2 +(△a2) 2 +(△b2) 2 ] 1 / 2 ; △L2 represents the brightness / black and white difference in different phenological periods: △L2=L 后一时期物候 -L 前一时期物候 ; △a2 represents the red / green difference in different phenological periods: △a2=a 后一时期物候 -a 前一时期物候 ; △b2 represents the difference in yellow / blue in different phenological periods: △b2=b 后一时期物候 -b 前一时期物候 ; During the subsequent use of the monitoring system, the vegetation color parameters of each plot were extracted, and the phenological period was determined by cluster analysis based on the color parameters. Phenological changes were monitored by comparing the changes in the color difference values ​​△E1' between different sampling points on the photos of a certain texture plot taken and the color difference values ​​△E2' between the photos of a certain texture plot taken and the last photo of the same germplasm taken. △E1' represents the color difference values ​​between different sampling points on the photos of a certain texture plot taken during the subsequent use of the monitoring system; △E2' represents the color difference values ​​between the photos of a certain texture plot taken and the last photo of the same germplasm taken during the subsequent use of the monitoring system.

2. The method for monitoring phenology of a woody ornamental plant germplasm resource bank according to claim 1, characterized in that: In step 1, an aerial image of the germplasm resource bank is obtained by vertically shooting with a fixed height using a drone. The image is ortho-mosaiced using the drone's built-in remote sensing image processing system, and then a reflectance image is obtained after radiation correction.

3. The method for monitoring phenology of a woody ornamental plant germplasm resource bank according to claim 1, characterized in that: In step 2, manual drawing is used to segment the boundaries of the subdistrict. The Quantum GIS vector operation open source library is used to add the orthophoto map as the base map, and then a new polygon feature layer is created. Then, each polygon feature is drawn according to the land parcel boundary of the resource library on the base map. One land parcel is one polygon feature, and one layer contains multiple polygon features.

4. The method for monitoring phenology of a woody ornamental plant germplasm resource bank according to claim 1, wherein: In step 3, different fixed time intervals are set for different phenological periods according to the characteristics of the tree species. Flowering tree species should be photographed at intervals of 1 day or half a day during their flowering phenological period, and the time intervals should be set to several days during non-important phenological periods.

5. The method for monitoring phenology of a woody ornamental plant germplasm resource bank according to claim 1, characterized in that: If △E1'>△E 1max This indicates that there are inconsistencies in phenological periods or alternate years or abnormal growth of flowers and fruits among different plants of the germplasm, and a prompt message is generated and sent to the investigator's communication device; If △E2' ≈ △E 2min , indicating that the phenological change is in a critical state, and a prompt message is generated and sent to the investigator's communication device; △E2' >△E 2min , indicating that the phenological stage spans a large distance, sending a warning message; △E 1max Indicates the maximum color difference between different sampling points on a certain texture patch photo in the phenology database; △E 2min Indicates the minimum color difference between adjacent phenological photos of a certain quality in the phenological database.

Images