Citrus orchard image data acquisition and labeling method

Abstract

The invention discloses a citrus orchard image data acquisition and annotation method, which relates to the technical field of image acquisition, and further comprises a factor analysis module, a default module, an image data standardization module, an image database module, an abnormal image analysis module, an abnormal root analysis module and a correction module. The system has the advantages that the quality of the acquired citrus orchard image can be obviously improved through the default module, the state and details of the fruit tree can be observed more clearly, the weather state information at the current acquisition time can be acquired by using the image data standardization module, the data is adjusted according to the image definition, and parameter adjustment data is generated, so that the accuracy of the system is improved. The reliability and the stability of image acquisition are improved, excessive influence of external factors such as weather change is avoided, the reason of the abnormal condition of the fruit tree can be analyzed according to the abnormal condition information of the fruit tree by utilizing the abnormal root analysis module, and the efficiency and the effect of problem solving are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image acquisition, and in particular to a method for acquiring and annotating image data of a citrus orchard. Background Art

[0002] Citrus is divided into tangerine and mandarin. The two have similar characteristics. The common point is that the peel is loose and easy to peel. The capsule is easy to separate, so it is also called wide-skinned citrus. It is an important tree species among citrus in China and even in Asia. Physiologically, tangerine is more primitive than tangerine. Mandarin is slightly more cold-resistant than tangerine. The wild characteristics and vitality of mandarin are stronger. It often grows in uninhabited areas. If the extreme types of tangerine and mandarin are compared, the difference is obvious. The fruit, peel and juice of citrus can be used as medicine, and can be used in daily necessities, packaging and other industries, so citrus will be planted on a large scale;

[0003] When using the common methods for collecting and labeling image data in citrus orchards, it is difficult to clearly observe the status and details of the fruit trees, which affects the accurate judgment of the growth of the fruit trees. It is also inconvenient to adjust the acquisition parameters of the image collector in real time according to the weather conditions, making image acquisition susceptible to external factors such as weather changes. Therefore, we propose a method for collecting and labeling image data in citrus orchards. Summary of the invention

[0004] The object of the present invention is to provide a method for collecting and labeling citrus orchard image data.

[0005] In order to solve the problems raised in the above background technology, the present invention provides the following technical solutions: a citrus orchard image data acquisition system, comprising an image data acquisition module, the image data acquisition module can collect image information of a citrus orchard, and also comprising a factor analysis module, a default module, an image data standardization module, an image database module and an abnormal image analysis module;

[0006] The default module can extract the image information of the citrus orchard, analyze the clarity of the image information of the citrus orchard, and determine the default quality of the image information in conjunction with the user to generate a default quality value;

[0007] The factor analysis module obtains the factors that affect the image clarity, analyzes the specific impact information of different factors on the image clarity, and generates image clarity adjustment data;

[0008] The image data standardization module obtains the time of currently collecting the image information of the citrus orchard, analyzes the weather conditions at the current collection time, generates parameter adjustment data according to the image clarity adjustment data, and then transmits the parameter adjustment data to the image collector, controls the image collector to collect the image information of the citrus orchard, and generates real-time image information;

[0009] The image database module obtains the type information of the fruit trees in the citrus orchard, and establishes an Internet unit and a custom unit. The Internet unit and the custom unit are used to analyze the image data of the citrus orchard under normal conditions, generate normal image information, and simultaneously obtain the appearance change information of abnormal conditions of different fruit trees, and generate abnormal change data;

[0010] The abnormal image annotation module extracts real-time image information, compares the real-time image information with the normal image information, checks the difference between the real-time image information and the normal image information, and uses the abnormal change data to determine the abnormal conditions of the fruit trees, and then marks the locations where the abnormal conditions occur.

[0011] As a further solution of the present invention: after extracting the image information of the citrus orchard, the default module will establish a collection unit and a display unit, then set an extraction threshold, use ImageJ to calculate the clarity of the citrus orchard image information, and after obtaining the clarity of the citrus orchard image information, it will be determined whether the clarity of the citrus orchard image information exceeds the extraction threshold, and the extraction threshold is set as T 阈值 , assuming that the clarity of the citrus orchard image information is Q 清晰度 ;

[0012] When Q 清晰度 <T 阈值 When the image information of the corresponding citrus orchard is recorded in the acquisition unit;

[0013] When Q 清晰度 ≥T 阈值 When the user determines the default quality of the image information, the user needs to confirm it according to the citrus orchard image information recorded in the display unit.

[0014] As a further solution of the present invention: after obtaining the factors affecting the image clarity, the factor analysis module will establish a rainy database, a normal database and a sunny database according to the weather conditions. When generating image clarity adjustment data, the image clarity adjustment data will be divided according to the weather status information to generate sub-adjustment data, and then the sub-adjustment data will be recorded in the corresponding rainy database, normal database and sunny database.

[0015] As a further solution of the present invention: after analyzing the weather status information at the current acquisition time, the image data standardization module will extract corresponding sub-adjustment data from the factor analysis module according to the analyzed weather status. After generating real-time image information, the clarity of the real-time image information will be calculated using ImageJ, and the difference between the clarity of the real-time image information and the default quality value will be analyzed. Then, the extracted sub-adjustment data will be used to analyze a first adjustment scheme that makes the clarity of the real-time image information reach the default quality value. Suppose various parameters of the real-time image information are G1, G2, G3, ..., G C , suppose the various parameter information of the sub-adjustment data are T1, T2, T3, ..., T C , assuming that the parameter information corresponding to various parameters of the real-time image information in the sub-adjustment data is S1, S2, S3, ..., S C ;

[0016] (S1, S2, S3, ..., S C )

[0017] =(G1, G2, G3, ..., G C )∩(T1, T2, T3, ..., T C )

[0018] According to the above formula, the parameter information corresponding to various parameters of the real-time image information in the sub-adjustment data can be calculated. Suppose the first-order adjustment scheme information is T F ;

[0019] T F =(G1-S1, G2-S2, G3-S3,...,G C -S C )

[0020] According to the above formula, the first adjustment scheme information can be calculated, and then the acquisition parameters of the image collector are adjusted again according to the first adjustment scheme, and the real-time image information is acquired again, and the clarity of the real-time image information is analyzed again to see whether it is higher than the default quality value;

[0021] When the clarity of the real-time image information is higher than the default quality value, the corresponding real-time image information is transmitted to the abnormal image annotation module;

[0022] When the clarity of the real-time image information is lower than the default quality value, the method for obtaining the first adjustment scheme information is repeated to obtain the second adjustment scheme information, and the acquisition parameters of the image collector are adjusted again according to the second adjustment scheme to obtain the real-time image information until the clarity of the real-time image information is higher than the default quality value.

[0023] As a further solution of the present invention: the Internet unit in the image database module can automatically access the Internet with fruit tree species information, establish folders according to fruit tree species, and after generating abnormal change data of different fruit tree species, the corresponding abnormal change data will be recorded in the corresponding folders, and at the same time, feature extraction is performed on the abnormal change data to obtain abnormal feature information, and the occurrence time information and state change information of different abnormal change data are obtained. Then, a line change graph is drawn according to the occurrence time information and state change information of different abnormal change data, and the line change graph and the abnormal feature information are marked on the corresponding folders.

[0024] As a further solution of the present invention: after acquiring the real-time image information, the abnormal image annotation module will extract the corresponding normal image information from the image database module according to the fruit tree species data in the real-time image information. When the corresponding normal image information cannot be retrieved from the image database module, a retrieval failure signal will be sent to the user. Suppose the species information of the real-time image information is Z. 种类 , assuming that the types of normal image information recorded by the image database module are T1, T2, T3, ..., T S ;

[0025] when When the search fails, a retrieval failure signal will be sent to the user;

[0026] when When the normal image information is retrieved, it will be displayed to the user;

[0027] Then the normal image information is compared with the real-time image information to obtain the difference information between the real-time image information and the normal image information, and the difference information is subjected to feature extraction to obtain the real-time difference data. Let the real-time difference data be C T , let the abnormal feature information of different folders be Y T , let the number of real-time difference data be N, let the abnormal compliance value be Y 符合值 , let the extracted value be T 提取值 ;

[0028]

[0029] According to the above formula, the abnormal compliance value can be calculated. After the first abnormal compliance value is calculated, it will directly affect T 提取值 To cover, the subsequent calculation of the abnormal compliance value needs to be consistent with T 提取值 Make comparisons;

[0030] When Y 符合值 >T 提取值 When 符合值 To T 提取值 To cover;

[0031] When Y 符合值 <T 提取值 When 符合值 delete;

[0032] When Y 符合值 =T 提取值 When 符合值 Marked in T 提取值 After the generation of the mark, the subsequent calculation of Y 符合值 >T 提取值 When Y is calculated 符合值 Will directly T 提取值 The T is output after all the abnormal compliance values ​​are calculated. 提取值 , and extract the 提取值 The corresponding abnormal feature information is framed and marked in the real-time image information, and the corresponding abnormal feature information is recorded on one side of the frame mark.

[0033] As a further solution of the present invention: after extracting the abnormal situation information of the fruit trees, the abnormal root cause analysis module can use the smart planting and farming to determine the cause of the abnormal situation of the fruit trees. The correction module can receive the information processed by the abnormal root cause analysis module. After determining the cause of the abnormal situation of the fruit trees, it will use the smart planting and farming to generate mitigation plans and solutions, and split the mitigation plans and solutions to generate sub-plan information. The cost of the mitigation plans and solutions is calculated according to the sub-plan information. Let the funds required for the sub-plan information be Z F , let the number of sub-scheme information be G, let the funds required for the same item information used in different sub-scheme information be X W , let the number of the same item information used in different sub-plans be H, let the plan capital cost be F 成本 ;

[0034]

[0035] According to the above formula, the financial cost of the plan can be calculated, and the time used to execute the plan information can be analyzed to calculate the time to execute the plan.

[0036] In addition, the present invention also provides a method for annotating a citrus orchard image, comprising the following steps:

[0037] S100, collecting image information of the citrus orchard, calculating the clarity of the image information of the citrus orchard using ImageJ, and generating a default quality value;

[0038] S200, obtaining factors affecting image clarity, analyzing specific information on the effects of different factors on image clarity, generating image clarity adjustment data, establishing a database, dividing the image clarity adjustment data, generating sub-adjustment data, and recording the sub-adjustment data into a corresponding database;

[0039] S300, analyzing the weather conditions at the current acquisition time, generating parameter adjustment data, adjusting the parameters of the image collector, acquiring image information of the citrus orchard, generating real-time image information, extracting corresponding sub-adjustment data, analyzing the difference between the clarity of the real-time image information and the default quality value, and generating an adjustment plan until the clarity of the real-time image information is higher than the default quality value;

[0040] S400, obtaining the types of fruit trees in the citrus orchard, analyzing the citrus orchard image data under normal conditions, generating normal image information, obtaining appearance change information of different fruit trees in abnormal conditions, generating abnormal change data, performing feature extraction on the abnormal change data, obtaining abnormal feature information, and drawing a broken line change graph;

[0041] S500, checking the difference between the real-time image information and the normal image information, using the abnormal change data to determine the abnormal situation of the fruit tree, selecting the position where the abnormal situation occurs in the real-time image information, extracting features from the difference information, obtaining real-time difference data, calculating the abnormal compliance value, and extracting corresponding abnormal feature information;

[0042] S600, using Smart Farmer and Farmer Identification to determine the cause of abnormal conditions in fruit trees, generate mitigation plans and solutions, and calculate the time cost information and financial cost information of the mitigation plans and solutions.

[0043] By adopting the above technical solution, compared with the prior art, the beneficial effects of the present invention are:

[0044] 1. The present invention can significantly improve the quality of the collected citrus orchard images through the default module, which is helpful to observe the status and details of the fruit trees more clearly. The image data standardization module can obtain the weather status information of the current collection time, and generate parameter adjustment data according to the image clarity adjustment data, which increases the reliability and stability of image collection and is not excessively affected by external factors such as weather changes. The abnormal root cause analysis module can analyze the causes of abnormal conditions in fruit trees according to the abnormal condition information of fruit trees, so that solutions can be provided in time, improving the efficiency and effect of problem solving;

[0045] 2. The present invention can effectively filter out blurred images through the default module, improve the efficiency and accuracy of subsequent analysis and processing, and allow users to participate in the quality confirmation process, ensuring the rationality and applicability of image quality standards. The factor analysis module can quickly and accurately extract the corresponding sub-adjustment data, greatly improving the efficiency and accuracy of image adjustment, and adapting to the image acquisition requirements under different weather conditions. The image data standardization module can maximize the clarity and quality of the image, providing a high-quality image foundation for subsequent abnormal image annotation and analysis;

[0046] 3. The present invention helps to classify and manage abnormal conditions of different fruit trees through the image database module, better understand and grasp the evolution process of abnormal conditions of fruit trees, improve the efficiency of image comparison, and provide convenience for quickly judging whether fruit trees are abnormal. The abnormal image annotation module can accurately judge the abnormal conditions of fruit trees, provide users with a clear abnormal position and feature display, and facilitate users to quickly locate and deal with problems. The use of the abnormal root cause analysis module helps users choose the most appropriate processing method according to actual conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Schematic diagram of the system flow in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The specific embodiments of the present invention will be further described below in conjunction with the accompanying drawings. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0049] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0050] Embodiment 1:

[0051] Therefore, in order to effectively solve the above problems, the present application proposes a method for collecting and annotating citrus orchard image data, as shown in the accompanying drawings of the specification. Figure 1 As shown, it includes an image data acquisition module, which can collect image information of a citrus orchard, and also includes a factor analysis module, a default module, an image data standardization module, an image database module, and an abnormal image analysis module;

[0052] The default module can extract the image information of the citrus orchard, analyze the clarity of the image information of the citrus orchard, and determine the default quality of the image information in conjunction with the user to generate a default quality value;

[0053] The factor analysis module obtains the factors that affect the image clarity, analyzes the specific impact information of different factors on the image clarity, and generates image clarity adjustment data;

[0054] The image data standardization module obtains the time of currently collecting the image information of the citrus orchard, analyzes the weather conditions at the current collection time, generates parameter adjustment data according to the image clarity adjustment data, and then transmits the parameter adjustment data to the image collector, controls the image collector to collect the image information of the citrus orchard, and generates real-time image information;

[0055] Predict future weather trends, develop standard weather units, access default modules, analyze weather status information that can capture default quality values, calculate the weather status information and match it with the weather status information in future weather trends, mark the best matching position within the set time, and remind the user;

[0056] The image database module obtains the type information of the fruit trees in the citrus orchard, and establishes an Internet unit and a custom unit. The Internet unit and the custom unit are used to analyze the image data of the citrus orchard under normal conditions, generate normal image information, and simultaneously obtain the appearance change information of abnormal conditions of different fruit trees, and generate abnormal change data;

[0057] Classifying fruit trees by their varieties, growth stages, pests and diseases, etc., can help the abnormal image annotation module to more accurately extract normal image information and determine abnormal conditions of fruit trees.

[0058] The abnormal image annotation module extracts real-time image information, compares the real-time image information with the normal image information, checks the difference between the real-time image information and the normal image information, and uses the abnormal change data to determine the abnormal situation of the fruit tree, and then marks the location where the abnormal situation occurs;

[0059] Establish a refinement unit. When an abnormal situation occurs, the image collector will be controlled to further zoom in on the selected position and adjust the acquisition parameters of the image collector again to make the acquired image more detailed and capture the details of the abnormal situation.

[0060] It also includes an abnormality root cause analysis module and a correction module;

[0061] The abnormal root cause analysis module can extract the abnormal situation information of the fruit trees, analyze the causes of the abnormal situation of the fruit trees based on the abnormal situation information of the fruit trees, and generate abnormal traceability data;

[0062] The correction module analyzes the methods of mitigating and resolving abnormal situations based on the abnormality tracing data, generates mitigation plans and solutions, and then analyzes the time and financial costs required for different mitigation plans and solutions;

[0063] Specific workflow: collect image information of citrus orchards, analyze the clarity of citrus orchard image information, determine the default quality of image information, generate default quality values, analyze the specific impact of different factors on image clarity, generate image clarity adjustment data, obtain the current time of collecting citrus orchard image information, generate parameter adjustment data according to the image clarity adjustment data, control the image collector to collect image information of citrus orchards, generate real-time image information, analyze citrus orchard image data under normal conditions, generate normal image information, synchronously obtain appearance change information of abnormal conditions of different fruit trees, generate abnormal change data, view the difference between real-time image information and normal image information, judge the abnormal conditions of fruit trees, select the locations where abnormal conditions occur, analyze the causes of abnormal conditions of fruit trees according to the abnormal condition information of fruit trees, generate abnormal traceability data, and analyze the time cost and financial cost required for different mitigation plans and different solutions;

[0064] Furthermore, the default module can significantly improve the quality of the collected citrus orchard images, which helps to observe the status and details of the fruit trees more clearly. The image data standardization module can obtain the weather status information at the current collection time, and generate parameter adjustment data based on the image clarity adjustment data, which increases the reliability and stability of image acquisition and is not overly affected by external factors such as weather changes. The abnormal root cause analysis module can analyze the causes of abnormal conditions in fruit trees based on the abnormal condition information of the fruit trees, so that solutions can be found in a timely manner, improving the efficiency and effectiveness of problem solving.

[0065] Embodiment 2:

[0066] Based on the first embodiment, as shown in the accompanying drawings of the specification Figure 1 As shown in the figure, after extracting the image information of the citrus orchard, the default module will establish a collection unit and a display unit, then set the extraction threshold, and use ImageJ to calculate the clarity of the citrus orchard image information. After obtaining the clarity of the citrus orchard image information, it will be determined whether the clarity of the citrus orchard image information exceeds the extraction threshold. The extraction threshold is set to

[0067] T 阈值 , assuming that the clarity of the citrus orchard image information is Q 清晰度 ;

[0068] When Q 清晰度 <T 阈值 When the image information of the corresponding citrus orchard is recorded in the acquisition unit;

[0069] When Q 清晰度 ≥T 阈值When the user determines the default quality of the image information, the user needs to confirm it according to the citrus orchard image information recorded in the display unit;

[0070] After obtaining the factors that affect the image clarity, the factor analysis module will establish a cloudy and rainy database, a normal database, and a sunny database according to the weather conditions. When generating image clarity adjustment data, the image clarity adjustment data will be divided according to the weather condition information to generate sub-adjustment data, and then the sub-adjustment data will be recorded in the corresponding cloudy and rainy database, normal database, and sunny database;

[0071] After analyzing the weather status information at the current acquisition time, the image data standardization module will extract the corresponding sub-adjustment data from the factor analysis module according to the analyzed weather status. After generating real-time image information, it will use ImageJ to calculate the clarity of the real-time image information, analyze the difference between the clarity of the real-time image information and the default quality value, and then use the extracted sub-adjustment data to analyze a first-level adjustment scheme that makes the clarity of the real-time image information reach the default quality value. Suppose the various parameters of the real-time image information are G1, G2, G3, ..., G C , suppose the various parameter information of the sub-adjustment data are T1, T2, T3, ..., T C , assuming that the parameter information corresponding to various parameters of the real-time image information in the sub-adjustment data is S1, S2, S3, ..., S C ;

[0072] (S1, S2, S3, ..., S C )

[0073] =(G1, G2, G3, ..., G C )∩(T1, T2, T3, ..., T C )

[0074] According to the above formula, the parameter information corresponding to various parameters of the real-time image information in the sub-adjustment data can be calculated. Suppose the first-order adjustment scheme information is T F ;

[0075] T F =(G1-S1, G2-S2, G3-S3,...,G C -S C )

[0076] According to the above formula, the first adjustment scheme information can be calculated, and then the acquisition parameters of the image collector are adjusted again according to the first adjustment scheme, and the real-time image information is acquired again, and the clarity of the real-time image information is analyzed again to see whether it is higher than the default quality value;

[0077] When the clarity of the real-time image information is higher than the default quality value, the corresponding real-time image information is transmitted to the abnormal image annotation module;

[0078] When the clarity of the real-time image information is lower than the default quality value, the method for obtaining the first adjustment scheme information is repeated to obtain the second adjustment scheme information, and the acquisition parameters of the image collector are adjusted again according to the second adjustment scheme to obtain the real-time image information until the clarity of the real-time image information is higher than the default quality value;

[0079] The parameter unit is adjusted slightly, and then the change in the clarity of the real-time image information is calculated, and then the parameter adjustment degree of the clarity of the real-time image information higher than the default quality value is calculated, and the parameter information of the slight adjustment is set as T C , let the corresponding real-time image information clarity change information be B C , assuming the clarity of the current real-time image information is D C , set the default mass value to M C , let different parameter adjustment information be C 调整 ;

[0080]

[0081] According to the above formula, different parameter adjustment information can be calculated. When all parameter adjustment information is calculated, adjustment plan information can be generated.

[0082] Specific workflow: establish a collection unit and a display unit, then set an extraction threshold, use ImageJ to calculate the clarity of the citrus orchard image information, determine whether the clarity of the citrus orchard image information exceeds the extraction threshold, establish a rainy database, a normal database and a sunny database according to the weather conditions, divide the image clarity adjustment data according to the weather status information, generate sub-adjustment data, and record them in the corresponding database, extract the corresponding sub-adjustment data from the factor analysis module according to the analyzed weather conditions, use ImageJ to calculate the clarity of the real-time image information, analyze the difference between the clarity of the real-time image information and the default quality value, calculate the parameter information corresponding to various parameters of the real-time image information in the sub-adjustment data, analyze a first-level adjustment plan, adjust the acquisition parameters of the image collector again according to the first-level adjustment plan, and obtain the real-time image information again, and analyze whether the clarity of the real-time image information is higher than the default quality value. When the clarity of the real-time image information is lower than the default quality value, repeat the method for obtaining the first-level adjustment plan information to obtain the second-level adjustment plan information, and adjust the acquisition parameters of the image collector again according to the second-level adjustment plan to obtain the real-time image information until the clarity of the real-time image information is higher than the default quality value;

[0083] Furthermore, the default module can effectively filter out blurred images, improve the efficiency and accuracy of subsequent analysis and processing, and allow users to participate in the quality confirmation process, ensuring the rationality and applicability of image quality standards. The factor analysis module can quickly and accurately extract the corresponding sub-adjustment data, greatly improving the efficiency and accuracy of image adjustment and adapting to image acquisition needs under different weather conditions. The image data standardization module can maximize the clarity and quality of the image, providing a high-quality image foundation for subsequent abnormal image annotation and analysis.

[0084] Embodiment three:

[0085] Based on the second embodiment, as shown in the accompanying drawings of the specification Figure 1 As shown, the Internet unit in the image database module can automatically access the Internet with fruit tree species information, establish folders according to fruit tree species, and after generating abnormal change data of different fruit tree species, the corresponding abnormal change data will be recorded in the corresponding folders, and at the same time, feature extraction is performed on the abnormal change data to obtain abnormal feature information, and at the same time, the occurrence time information and state change information of different abnormal change data are obtained, and then a broken line change graph is drawn according to the occurrence time information and state change information of different abnormal change data, and the broken line change graph and the abnormal feature information are marked on the corresponding folders;

[0086] Collect other information related to fruit tree growth, such as soil type, climate data, surrounding ecological environment, etc., and combine this information with abnormal change data to conduct more comprehensive feature extraction and analysis to better understand the causes of fruit tree abnormalities;

[0087] Share the data in the image database module with other citrus orchard managers or agricultural research institutions to promote information exchange and cooperation. Through the sharing platform, more abnormal change cases and solutions can be collected, enriching the content of the database and improving the entire industry's ability to respond to abnormal situations in citrus orchards;

[0088] After obtaining the real-time image information, the abnormal image annotation module will extract the corresponding normal image information from the image database module according to the fruit tree species data in the real-time image information. When the corresponding normal image information cannot be retrieved from the image database module, a retrieval failure signal will be sent to the user. Suppose the species information of the real-time image information is Z. 种类 , assuming that the types of normal image information recorded by the image database module are T1, T2, T3, ..., T S ;

[0089] when When the search fails, a retrieval failure signal will be sent to the user;

[0090] when When the normal image information is retrieved, it will be displayed to the user;

[0091] Then the normal image information is compared with the real-time image information to obtain the difference information between the real-time image information and the normal image information, and the difference information is subjected to feature extraction to obtain the real-time difference data. Let the real-time difference data be C T , let the abnormal feature information of different folders be Y T , let the number of real-time difference data be N, let the abnormal compliance value be Y 符合值 , let the extracted value be T 提取值 ;

[0092]

[0093] According to the above formula, the abnormal compliance value can be calculated. After the first abnormal compliance value is calculated, it will directly affect T 提取值 To cover, the subsequent calculation of the abnormal compliance value needs to be consistent with T 提取值 Make comparisons;

[0094] When Y 符合值 >T 提取值 When 符合值 To T 提取值 To cover;

[0095] When Y 符合值 <T 提取值 When 符合值 delete;

[0096] When Y 符合值 =T 提取值 When 符合值 Marked in T 提取值 After the generation of the mark, the subsequent calculation of Y 符合值 >T 提取值 When Y is calculated 符合值 Will directly T 提取值 The T is output after all the abnormal compliance values ​​are calculated. 提取值 , and extract the 提取值 The corresponding abnormal feature information is framed and marked in the real-time image information, and the corresponding abnormal feature information is recorded on one side of the frame mark;

[0097] After extracting the abnormal situation information of the fruit trees, the abnormal root cause analysis module can use the smart farming and farmer identification to judge the cause of the abnormal situation of the fruit trees. The correction module can receive the information processed by the abnormal root cause analysis module. After judging the cause of the abnormal situation of the fruit trees, it will use the smart farming and farmer identification to generate mitigation plans and solutions, and split the mitigation plans and solutions to generate sub-plan information. The cost of the mitigation plans and solutions is calculated based on the sub-plan information. Suppose the funds required for the sub-plan information is Z F , let the number of sub-scheme information be G, let the funds required for the same item information used in different sub-scheme information be X W , let the number of the same item information used in different sub-plans be H, let the plan capital cost be F 成本 ;

[0098]

[0099] According to the above formula, the financial cost of the plan can be calculated, and the time used to execute the plan information can be analyzed to calculate the time to execute the plan;

[0100] Specific workflow: Create folders according to the types of fruit trees, record the corresponding abnormal change data in the corresponding folders, extract features from the abnormal change data, obtain the occurrence time information and state change information of different abnormal change data, draw a broken line change graph, extract the corresponding normal image information from the image database module according to the fruit tree type data in the real-time image information, and when the corresponding normal image information cannot be retrieved from the image database module, a retrieval failure signal will be sent to the user, and the normal image information will be compared with the real-time image information to obtain the difference information between the real-time image information and the normal image information, calculate the abnormal compliance value, extract the abnormal feature information corresponding to the extracted value, and record it on one side of the box mark, use Smart Farmer and Farmer Identification to judge the cause of the abnormal situation of the fruit tree, and use Smart Farmer and Farmer Identification to generate mitigation plans and solutions, split the mitigation plans and solutions, generate sub-plan information, calculate the financial cost of the plan and the time to execute the plan, and display them;

[0101] Furthermore, the image database module helps to classify and manage abnormal conditions of different fruit trees, better understand and grasp the evolution process of abnormal conditions of fruit trees, improve the efficiency of image comparison, and facilitate rapid judgment of whether fruit trees are abnormal. The abnormal image annotation module can accurately judge the abnormal conditions of fruit trees, and provide users with a clear display of abnormal locations and features, making it convenient for users to quickly locate and deal with problems. The abnormal root cause analysis module helps users choose the most appropriate treatment method based on actual conditions.

[0102] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A citrus orchard image data acquisition system, comprising an image data acquisition module, wherein the image data acquisition module can acquire image information of a citrus orchard, and wherein: It also includes a factor analysis module, a default module, an image data standardization module, an image database module, and an abnormal image analysis module; The default module can extract the image information of the citrus orchard, analyze the clarity of the image information of the citrus orchard, and determine the default quality of the image information in conjunction with the user to generate a default quality value; The factor analysis module obtains the factors that affect the image clarity, analyzes the specific impact information of different factors on the image clarity, and generates image clarity adjustment data; The image data standardization module obtains the time of currently collecting the image information of the citrus orchard, analyzes the weather conditions at the current collection time, generates parameter adjustment data according to the image clarity adjustment data, and then transmits the parameter adjustment data to the image collector, controls the image collector to collect the image information of the citrus orchard, and generates real-time image information; The image database module obtains the type information of the fruit trees in the citrus orchard, and establishes an Internet unit and a custom unit. The Internet unit and the custom unit are used to analyze the image data of the citrus orchard under normal conditions, generate normal image information, and simultaneously obtain the appearance change information of abnormal conditions of different fruit trees, and generate abnormal change data; The abnormal image annotation module extracts real-time image information, compares the real-time image information with the normal image information, checks the difference between the real-time image information and the normal image information, and uses the abnormal change data to determine the abnormal conditions of the fruit trees, and then marks the locations where the abnormal conditions occur.

2. The citrus orchard image data acquisition system according to claim 1, characterized in that: It also includes an abnormality root cause analysis module and a correction module; The abnormal root cause analysis module can extract the abnormal situation information of the fruit trees, analyze the causes of the abnormal situation of the fruit trees based on the abnormal situation information of the fruit trees, and generate abnormal traceability data; The correction module analyzes methods to mitigate and resolve abnormal situations based on the abnormality tracing data, generates mitigation plans and solutions, and then analyzes the time and financial costs required for different mitigation plans and solutions.

3. The citrus orchard image data acquisition system according to claim 1, characterized in that: After extracting the image information of the citrus orchard, the default module will establish a collection unit and a display unit, then set an extraction threshold, use ImageJ to calculate the clarity of the citrus orchard image information, and after obtaining the clarity of the citrus orchard image information, it will be determined whether the clarity of the citrus orchard image information exceeds the extraction threshold. The extraction threshold is set as T 阅值 , assuming that the clarity of the citrus orchard image information is Q 清晰度 ; When Q 清晰度 <T 阈值 When the image information of the corresponding citrus orchard is recorded in the acquisition unit; When Q 清晰度 ≥T 阈值 When the user determines the default quality of the image information, the user needs to confirm it according to the citrus orchard image information recorded in the display unit.

4. The citrus orchard image data acquisition system according to claim 1, characterized in that: After obtaining the factors that affect the image clarity, the factor analysis module will establish a rainy database, a normal database, and a sunny database according to the weather conditions. When generating image clarity adjustment data, the image clarity adjustment data will be divided according to the weather status information to generate sub-adjustment data, and then the sub-adjustment data will be recorded in the corresponding rainy database, normal database, and sunny database.

5. The citrus orchard image data acquisition system according to claim 1, characterized in that: After analyzing the weather status information at the current acquisition time, the image data standardization module extracts corresponding sub-adjustment data from the factor analysis module according to the analyzed weather status. After generating real-time image information, the image data standardization module calculates the clarity of the real-time image information using ImageJ, analyzes the difference between the clarity of the real-time image information and the default quality value, and then uses the extracted sub-adjustment data to analyze a first-order adjustment scheme that makes the clarity of the real-time image information reach the default quality value. Suppose various parameters of the real-time image information are G1, G2, G3, ..., G C , suppose the various parameter information of the sub-adjustment data are T1, T2, T3, ..., T C , assuming that the parameter information corresponding to various parameters of the real-time image information in the sub-adjustment data is S1, S2, S3, ..., S C ; <h2 style=";text-align:left;direction:ltr">(S1, S2, S3,…, S<h2 style=";text-align:left;direction:ltr"> C <h2 style=";text-align:left;direction:ltr"> )=(G1、G2、G3、......、G<h2 style=";text-align:left;direction:ltr"> C <h2 style=";text-align:left;direction:ltr"> )∩(T1、T2、T3、......、T<h2 style=";text-align:left;direction:ltr"> C <h2 style=";text-align:left;direction:ltr"> ) According to the above formula, the parameter information corresponding to various parameters of the real-time image information in the sub-adjustment data can be calculated. Suppose the first-order adjustment scheme information is T F ; T F =(G1-S1、G2-S2、G3-S3、……、G C -S C ) According to the above formula, the first adjustment scheme information can be calculated, and then the acquisition parameters of the image collector are adjusted again according to the first adjustment scheme, and the real-time image information is acquired again, and the clarity of the real-time image information is analyzed again to see whether it is higher than the default quality value; When the clarity of the real-time image information is higher than the default quality value, the corresponding real-time image information is transmitted to the abnormal image annotation module; When the clarity of the real-time image information is lower than the default quality value, the method for obtaining the first adjustment scheme information is repeated to obtain the second adjustment scheme information, and the acquisition parameters of the image collector are adjusted again according to the second adjustment scheme to obtain the real-time image information until the clarity of the real-time image information is higher than the default quality value.

6. The citrus orchard image data acquisition system according to claim 5, characterized in that: The Internet unit in the image database module can automatically access the Internet with fruit tree species information, establish folders according to fruit tree species, and after generating abnormal change data of different fruit tree species, record the corresponding abnormal change data in the corresponding folders, and perform feature extraction on the abnormal change data to obtain abnormal feature information, and obtain the occurrence time information and state change information of different abnormal change data, and then draw a line change graph according to the occurrence time information and state change information of different abnormal change data, and mark the line change graph and abnormal feature information on the corresponding folders.

7. A citrus orchard image data acquisition system according to claim 6, characterized in that: After acquiring the real-time image information, the abnormal image annotation module will extract the corresponding normal image information from the image database module according to the fruit tree species data in the real-time image information. When the corresponding normal image information cannot be retrieved from the image database module, a retrieval failure signal will be sent to the user. Suppose the species information of the real-time image information is Z. 种类 , assuming that the types of normal image information recorded by the image database module are T1, T2, T3, ..., T S ; when When the search fails, a retrieval failure signal will be sent to the user; when When the normal image information is retrieved, it will be displayed to the user; Then the normal image information is compared with the real-time image information to obtain the difference information between the real-time image information and the normal image information, and the difference information is subjected to feature extraction to obtain the real-time difference data. Let the real-time difference data be C T , let the abnormal feature information of different folders be Y T , let the number of real-time difference data be N, let the abnormal compliance value be Y 符号值 , let the extracted value be T 提取值 ; According to the above formula, the abnormal compliance value can be calculated. After the first abnormal compliance value is calculated, it will directly affect T 提取值 To cover, the subsequent calculation of the abnormal compliance value needs to be consistent with T 提取值 Make comparisons; When Y 符合值 >T 提取值 When 符合值 To T 提取值 To cover; When Y 符合值 <T 提取值 When 符合值 delete; When Y 符合值 =T 提取值 When 符合值 Marked in T 提取值 After the generation of the mark, the subsequent calculation of Y 符合值 >T 提取值 When Y is calculated 符合值 Will directly T 提取值 The T is output after all the abnormal compliance values ​​are calculated. 提取值 , and extract the 提取值 The corresponding abnormal feature information is framed and marked in the real-time image information, and the corresponding abnormal feature information is recorded on one side of the frame mark.

8. The citrus orchard image data acquisition system according to claim 2, characterized in that: After extracting the abnormal situation information of the fruit trees, the abnormal root cause analysis module can use the smart farming and farmer identification to judge the cause of the abnormal situation of the fruit trees. The correction module can receive the information processed by the abnormal root cause analysis module. After judging the cause of the abnormal situation of the fruit trees, it will use the smart farming and farmer identification to generate mitigation plans and solutions, and split the mitigation plans and solutions to generate sub-plan information. The cost of the mitigation plans and solutions is calculated according to the sub-plan information. Let the funds required for the sub-plan information be Z F , let the number of sub-scheme information be G, let the funds required for the same item information used in different sub-scheme information be X W , let the number of the same item information used in different sub-plans be H, let the plan capital cost be F 成本 ; According to the above formula, the financial cost of the plan can be calculated, and the time used to execute the plan information can be analyzed to calculate the time to execute the plan.

9. A method for annotating citrus orchard images applicable to the citrus orchard image data acquisition system of any one of claims 1 to 8, characterized in that: The following steps are involved: S100, collecting image information of the citrus orchard, calculating the clarity of the image information of the citrus orchard using imageJ, and generating a default quality value; S200, obtaining factors affecting image clarity, analyzing specific information on the effects of different factors on image clarity, generating image clarity adjustment data, establishing a database, dividing the image clarity adjustment data, generating sub-adjustment data, and recording the sub-adjustment data into a corresponding database; S300, analyzing the weather conditions at the current acquisition time, generating parameter adjustment data, adjusting the parameters of the image collector, acquiring image information of the citrus orchard, generating real-time image information, extracting corresponding sub-adjustment data, analyzing the difference between the clarity of the real-time image information and the default quality value, and generating an adjustment plan until the clarity of the real-time image information is higher than the default quality value; S400, obtaining the types of fruit trees in the citrus orchard, analyzing the citrus orchard image data under normal conditions, generating normal image information, obtaining appearance change information of different fruit trees in abnormal conditions, generating abnormal change data, performing feature extraction on the abnormal change data, obtaining abnormal feature information, and drawing a broken line change graph; S500, checking the difference between the real-time image information and the normal image information, using the abnormal change data to determine the abnormal situation of the fruit tree, selecting the position where the abnormal situation occurs in the real-time image information, extracting features from the difference information, obtaining real-time difference data, calculating the abnormal compliance value, and extracting corresponding abnormal feature information; S600, using Smart Farmer and Farmer Identification to determine the cause of abnormal conditions in fruit trees, generate mitigation plans and solutions, and calculate the time cost information and financial cost information of the mitigation plans and solutions.

Images