Intelligent detection system and method for neck elongation of citrus variety
Through the intelligent detection system that links high-definition cameras with LED light boards, combined with advanced image processing technology and environmental parameter acquisition, the problems of low efficiency and environmental adaptability in citrus neck length detection are solved, and high-precision and rapid fruit neck length measurement is achieved.
Patent Information
- Application Number
- CN202510812545.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-18
- Publication Date
- 2025-09-19
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing technologies are inefficient in detecting the neck length of citrus fruit, are easily interfered with by human factors, and have difficulty obtaining high-quality image information in complex environments, which affects the detection precision and accuracy.
It uses a high-definition camera linked with an LED light board, combined with image processing technologies such as median filtering, histogram equalization, HSV color space conversion and binary mask generation, combined with an environmental parameter acquisition module, and establishes a linear model through a data analysis module to screen key variables and achieve accurate measurement of fruit neck length.
It improves the accuracy and efficiency of fruit neck length measurement, reduces human errors, realizes non-contact rapid and accurate detection, enhances the accuracy and reliability of the test results, and adapts to complex and changing natural environments.
Smart Images
Figure HDA0005454324750000011 
Figure HDA0005454324750000021
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of intelligent sensing systems, and in particular to an intelligent detection system and method for citrus fruit neck elongation. Background Art
[0002] As one of the world's most important economic crops, citrus fruit variety improvement and quality control have always been core concerns in agricultural research and production practice. As a key indicator of citrus fruit morphology, neck length not only affects the appearance quality of the fruit but is also closely related to the fruit's growth condition, stress resistance, and ultimate commercial value. Traditionally, the measurement of citrus neck length relies on manual visual inspection or simple tool measurement. This method is inefficient and susceptible to human interference, making it difficult to achieve large-scale, high-precision detection. The rapid development of information technology, especially the widespread application of image processing and machine learning technologies, has provided new technical means for accurate detection in the agricultural field. However, existing technologies still face many challenges when applied to citrus neck elongation detection. First, the complex and variable citrus growth environment (such as light, temperature and humidity) significantly affects the quality of image acquisition, making it a major challenge to stably obtain high-quality image information under different environments. Second, factors such as occlusion between citrus fruit and branches and color similarity make accurate segmentation and recognition of the neck region more difficult. In addition, how to effectively combine environmental parameters and image information, deeply explore the intrinsic relationship between them and the length of the fruit neck, and then realize the intelligent prediction and regulation of the elongation of the fruit neck, is also a problem that needs to be solved urgently by current technology. Summary of the Invention
[0003] In view of this, the present invention addresses the deficiencies in the prior art and proposes an intelligent detection system and method for citrus fruit neck elongation, aiming to solve at least one of the problems raised in the above-mentioned background technology.
[0004] In a first aspect, the present invention provides an intelligent detection system for citrus fruit neck elongation, comprising: an image acquisition module configured to obtain image information of the citrus fruit neck to be detected, and process the image information to obtain the citrus fruit neck length;
[0005] An environmental parameter acquisition module is configured to acquire environmental parameters of the growth environment of the citrus to be tested;
[0006] The data analysis module is electrically connected to the environmental parameter acquisition module and the image acquisition module respectively, and is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length.
[0007] In some embodiments, when the image acquisition module is configured to acquire image information of the citrus neck to be detected, it includes:
[0008] The image acquisition module is also configured to obtain image information of the citrus fruit neck to be detected through a high-definition camera. An LED light board is set under each high-definition camera, and the LED light board is linked to the high-definition camera.
[0009] In some embodiments, when the image acquisition module processes the image information, it includes:
[0010] The image acquisition module is further configured to perform denoising on the image information. The denoising adopts a median filtering algorithm, the filter window size is set to 3x3, and the contrast of the denoised image information is compared with a preset contrast of the image acquisition module, wherein:
[0011] If the contrast is lower than a preset contrast of the image acquisition module, the image acquisition module improves the contrast of the image information after denoising by using histogram equalization;
[0012] If the contrast is greater than or equal to a preset contrast of the image acquisition module, the image acquisition module does not improve the contrast of the image information after denoising by using histogram equalization.
[0013] In some embodiments, when the image acquisition module processes the image information, it includes:
[0014] The image acquisition module is further configured to convert the image information from a default RGB color space to an HSV color space, and set an H channel threshold, an S channel threshold, and a V channel threshold, and generate a binary mask based on the H channel threshold, the S channel threshold, and the V channel threshold to obtain a binary image including the citrus neck area;
[0015] The image acquisition module is further configured to calculate the gray level co-occurrence matrix of the binary image and extract the corresponding texture contrast and texture correlation, wherein:
[0016] If the texture contrast of a certain area is lower than the preset texture contrast threshold, the area is judged as branch and leaf debris and removed from the foreground;
[0017] If the texture correlation of a certain area is greater than or equal to a preset texture correlation threshold, the area is determined to be the surface of the citrus neck and is retained as the foreground.
[0018] In some embodiments, when the image acquisition module obtains the neck length of a citrus variety, it includes:
[0019] The image acquisition module is further configured to use the Canny operator of the edge detection algorithm to extract the edge of the citrus fruit contour, set a threshold for edge detection, and use an ellipse fitting algorithm to fit the citrus fruit contour to obtain characteristic parameters of the citrus fruit;
[0020] The image acquisition module is further configured to convert the image information into an RGB color space and calculate the average R, G, and B values of the citrus fruit region as a color feature vector;
[0021] The image acquisition module is further configured to determine the position of the neck of the citrus fruit according to the shape characteristics and color characteristics of the citrus fruit, including:
[0022] By analyzing the color changes and texture differences in the top area of the citrus fruit after the ellipse fitting, the starting point of the fruit neck is found. The position of the fruit stalk is searched along the citrus fruit outline to determine the end point of the fruit neck. The pixel distance of the fruit neck in the image is calculated and converted into actual length according to the calibration parameters of the camera.
[0023] In some embodiments, when the environmental parameter acquisition module is configured to acquire environmental parameters of the growth environment of the citrus to be detected, it includes:
[0024] The environmental parameter acquisition module is also configured to set multiple temperature sensors, humidity sensors and light intensity sensors at the top and bottom of the environment in which the citrus is located, and obtain temperature data, humidity data and light intensity data of the environment in which the citrus is located in real time.
[0025] In some embodiments, when the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes:
[0026] The data analysis module is further configured to calculate the correlation coefficient between the environmental parameter and the length of the citrus neck, and the correlation coefficient between the image information and the length of the citrus neck.
[0027] In some embodiments, when the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes:
[0028] The data analysis module is further configured to establish a linear model between the citrus neck length and the environmental parameters and image information, and analyze the significance of each variable.
[0029] In some embodiments, when the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes:
[0030] The data analysis module is further configured to screen out key variables based on the feature importance of the linear model.
[0031] In a second aspect, the present invention provides an intelligent detection method for citrus fruit neck elongation, comprising the following steps:
[0032] S1. Use high-definition cameras to acquire image information of the citrus fruit necks to be inspected. An LED light panel is installed below each camera to provide linked lighting. The image is denoised using a 3x3 median filter algorithm. The contrast is compared with a preset value to determine whether histogram equalization is performed to improve the contrast.
[0033] S2, converting the processed image from RGB color space to HSV color space, setting H, S, and V channel thresholds to generate a binary mask to separate the area containing the citrus neck;
[0034] S3. Calculate the gray-level co-occurrence matrix of the binary image, extract texture contrast and correlation, and determine whether the area is a branch or leaf fragment or the surface of a citrus fruit neck based on a preset threshold, thereby accurately extracting the foreground;
[0035] S4. Use the Canny operator to perform edge detection and set a threshold. Use the ellipse fitting algorithm to obtain the characteristic parameters of the citrus fruit. Convert back to the RGB color space to calculate the average color value. Combine the shape and color features to determine the position of the fruit neck. Calculate the pixel distance and convert it into actual length according to the calibration parameters.
[0036] S5. Arrange multiple sensors at the top and bottom of the citrus growing environment to collect temperature, humidity, and light intensity data in real time;
[0037] S6. Calculate the correlation coefficients between environmental parameters and fruit neck length and the correlation coefficients between image information and citrus fruit neck length, establish a linear model to analyze the significance between the variables, and screen out the key variables affecting fruit neck length based on the model.
[0038] Compared with the existing technology, the beneficial effects of the present invention are: through the linkage lighting system of high-definition cameras and LED light panels, combined with advanced image processing technologies (such as median filtering, histogram equalization, HSV color space conversion and binary mask generation, etc.), it can effectively remove noise, enhance image contrast, and accurately segment the citrus neck area, greatly improving the accuracy and efficiency of neck length measurement. Compared with traditional manual measurement methods, human errors are reduced, and non-contact, fast and accurate detection is achieved. The system not only focuses on image information, but also integrates an environmental parameter acquisition module to monitor data such as temperature, humidity and light intensity of the citrus growth environment in real time. This design enables the system to fully consider the impact of environmental factors on fruit neck growth, improve the accuracy and reliability of detection results, and maintain stable detection performance even in complex and changeable natural environments. The data analysis module calculates the correlation coefficient between environmental parameters and fruit neck length, establishes a linear model, and analyzes the significance between each variable, which can deeply explore the intrinsic relationship between environmental parameters, image features and fruit neck length. Model-based feature importance screening can identify key variables affecting fruit neck length, provide a scientific basis for the selection of citrus varieties and optimization of cultivation management, and realize precision agricultural management.
[0039] The foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure.
[0040] Other features and aspects of the present disclosure will become more apparent from the following detailed description of exemplary embodiments with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0042] Figure 1 This is a functional block diagram of an intelligent detection system for citrus fruit neck elongation provided by an embodiment of the present invention;
[0043] Figure 2 Flowchart of the intelligent detection method for citrus fruit neck elongation provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0044] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0045] In the description of this application, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application.
[0046] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature specified as "first" or "second" may explicitly or implicitly include one or more of such features. Throughout this application, unless otherwise specified, "plurality" means two or more.
[0047] In the description of this application, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.
[0048] See Figure 1-2 As shown, the first embodiment: an intelligent detection system for citrus fruit neck elongation according to an embodiment of the present application, comprising:
[0049] An image acquisition module is configured to obtain image information of the neck of the citrus fruit to be detected, and process the image information to obtain the neck length of the citrus fruit variety;
[0050] An environmental parameter acquisition module is configured to acquire environmental parameters of the growth environment of the citrus to be tested;
[0051] The data analysis module is electrically connected to the environmental parameter acquisition module and the image acquisition module respectively, and is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length.
[0052] In some specific embodiments, when the image acquisition module is configured to acquire image information of the citrus neck to be detected, it includes:
[0053] The image acquisition module is also configured to obtain image information of the citrus fruit neck to be detected through a high-definition camera. An LED light board is set under each high-definition camera, and the LED light board is linked to the high-definition camera.
[0054] In some specific embodiments, when the image acquisition module processes the image information, it includes:
[0055] The image acquisition module is further configured to perform denoising on the image information. The denoising adopts a median filtering algorithm, the filter window size is set to 3x3, and the contrast of the denoised image information is compared with a preset contrast of the image acquisition module, wherein:
[0056] If the contrast is lower than a preset contrast of the image acquisition module, the image acquisition module improves the contrast of the image information after denoising by using histogram equalization;
[0057] If the contrast is greater than or equal to a preset contrast of the image acquisition module, the image acquisition module does not improve the contrast of the image information after denoising by using histogram equalization.
[0058] In some specific embodiments, when the image acquisition module processes the image information, it includes:
[0059] The image acquisition module is further configured to convert the image information from a default RGB color space to an HSV color space, and set an H channel threshold, an S channel threshold, and a V channel threshold, and generate a binary mask based on the H channel threshold, the S channel threshold, and the V channel threshold to obtain a binary image including the citrus neck area;
[0060] The image acquisition module is further configured to calculate the gray level co-occurrence matrix of the binary image and extract the corresponding texture contrast and texture correlation, wherein:
[0061] If the texture contrast of a certain area is lower than the preset texture contrast threshold, the area is judged as branch and leaf debris and removed from the foreground;
[0062] If the texture correlation of a certain area is greater than or equal to a preset texture correlation threshold, the area is determined to be the surface of the citrus neck and is retained as the foreground.
[0063] In some specific embodiments, when the image acquisition module obtains the neck length of a citrus variety, it includes:
[0064] The image acquisition module is further configured to use the Canny operator of the edge detection algorithm to extract the edge of the citrus fruit contour, set a threshold for edge detection, and use an ellipse fitting algorithm to fit the citrus fruit contour to obtain characteristic parameters of the citrus fruit;
[0065] The image acquisition module is further configured to convert the image information into an RGB color space and calculate the average R, G, and B values of the citrus fruit region as a color feature vector;
[0066] The image acquisition module is further configured to determine the position of the neck of the citrus fruit according to the shape characteristics and color characteristics of the citrus fruit, including:
[0067] By analyzing the color changes and texture differences in the top area of the citrus fruit after the ellipse fitting, the starting point of the fruit neck is found. The position of the fruit stalk is searched along the citrus fruit outline to determine the end point of the fruit neck. The pixel distance of the fruit neck in the image is calculated and converted into actual length according to the calibration parameters of the camera.
[0068] In some specific embodiments, when the environmental parameter acquisition module is configured to acquire the environmental parameters of the growth environment of the citrus to be detected, it includes:
[0069] The environmental parameter acquisition module is also configured to set multiple temperature sensors, humidity sensors and light intensity sensors at the top and bottom of the environment in which the citrus is located, and obtain temperature data, humidity data and light intensity data of the environment in which the citrus is located in real time.
[0070] In some specific embodiments, when the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes:
[0071] The data analysis module is further configured to calculate the correlation coefficient between the environmental parameter and the length of the citrus neck, and the correlation coefficient between the image information and the length of the citrus neck.
[0072] It should be understood that temperature, humidity, light intensity, and other data are extracted from the environmental parameter acquisition module. Citrus neck image feature data, including color features (such as RGB mean) and texture features (such as contrast and correlation of the gray-level co-occurrence matrix), are extracted from the image acquisition module. Measured neck length data is obtained from the image acquisition module. Data integrity is checked, missing values or outliers are removed, and environmental parameters, image features, and neck length data are aligned by sample to form a unified dataset.
[0073] Environmental parameters (such as temperature, humidity, and light intensity) and fruit neck length data were extracted from the dataset. Pearson correlation coefficients were calculated for each environmental parameter (such as temperature) and fruit neck length. The correlation coefficients between each environmental parameter and fruit neck length were calculated to determine the correlation between temperature, humidity, light intensity, and fruit neck length.
[0074] Image features (such as RGB mean, texture contrast, and texture correlation) and fruit neck length data were extracted from the dataset. Pearson correlation coefficients were calculated between each image feature (such as the R channel mean) and fruit neck length using the same method as above. The correlation coefficients between each image feature and fruit neck length were calculated to determine the correlation between color and texture features and fruit neck length.
[0075] Each correlation coefficient was tested for statistical significance using a t-test. Based on the t-statistic and significance level (e.g., 0.05), a critical value was found to determine whether the correlation coefficient was significant. If the correlation coefficient was significant (p < 0.05), the environmental parameter or image feature was significantly correlated with fruit neck length. If it was not significant, the parameter or feature was considered to have a minimal effect on fruit neck length.
[0076] A multivariate linear regression model was established, using fruit neck length as the dependent variable and environmental parameters and image features as independent variables. Through stepwise regression or feature importance screening, insignificant variables were removed, retaining the key variables with the greatest impact on fruit neck length. Correlation coefficients were calculated to clarify the extent to which environmental parameters and image features affect fruit neck length, providing a basis for precise management. Significance testing and feature screening were used to remove insignificant variables, simplify model complexity, and improve prediction accuracy. Based on correlation analysis and linear models, targeted management strategies (such as regulating temperature and light or optimizing image acquisition parameters) can be developed to promote fruit neck lengthening.
[0077] In some specific embodiments, when the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes:
[0078] The data analysis module is further configured to establish a linear model between the citrus neck length and the environmental parameters and image information, and analyze the significance of each variable.
[0079] In some specific embodiments, when the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes:
[0080] The data analysis module is further configured to screen out key variables based on the feature importance of the linear model.
[0081] It should be understood that high-definition cameras capture images of citrus fruit necks. LED panels are installed beneath each camera to provide uniform, shadow-free illumination, ensuring image clarity and consistency. A 3x3 median filter algorithm is used to remove image noise. The contrast of the denoised image is compared with a preset threshold to determine whether histogram equalization is performed to enhance contrast.
[0082] The image was converted from RGB to HSV color space, and thresholds were set on the H, S, and V channels to generate a binary mask to isolate the citrus neck region. The gray-level co-occurrence matrix of the binary image was calculated to extract texture contrast and correlation. Branch and leaf debris (low-contrast areas) were removed, while retaining the neck surface (high-correlation areas).
[0083] The Canny operator was used to extract the contour edge of the citrus fruit. The fruit shape parameters were obtained through the ellipse fitting algorithm. The fruit neck position was determined by combining the color features (RGB mean), and the fruit neck length was calculated.
[0084] The LED light panel and camera are linked to ensure uniform lighting, minimizing shadows and reflections. Median filtering and histogram equalization enhance image quality and adapt to varying lighting conditions. HSV color space conversion and texture analysis (gray-level co-occurrence matrix) effectively distinguish the fruit neck from branch and leaf fragments, avoiding complex background interference. Canny edge detection and ellipse fitting algorithms quickly locate the fruit neck. Combined with calibration parameters, pixel distance is converted to actual length, enabling non-contact, high-precision measurement.
[0085] Multiple temperature, humidity, and light intensity sensors are placed at the top and bottom of the citrus growing environment to collect real-time environmental data. Real-time temperature, humidity, and light intensity data provide a comprehensive basis for analyzing the impact of the environment on fruit neck growth. Combining environmental parameters with fruit neck length data can optimize management measures such as irrigation, fertilization, and light regulation to improve citrus quality.
[0086] In a second embodiment, an intelligent detection method for citrus fruit neck elongation according to an embodiment of the present application includes the following steps:
[0087] S1. Use high-definition cameras to acquire image information of the citrus fruit necks to be inspected. An LED light panel is installed below each camera to provide linked lighting. The image is denoised using a 3x3 median filter algorithm. The contrast is compared with a preset value to determine whether histogram equalization is performed to improve the contrast.
[0088] S2, converting the processed image from RGB color space to HSV color space, setting H, S, and V channel thresholds to generate a binary mask to separate the area containing the citrus neck;
[0089] S3. Calculate the gray-level co-occurrence matrix of the binary image, extract texture contrast and correlation, and determine whether the area is a branch or leaf fragment or the surface of a citrus fruit neck based on a preset threshold, thereby accurately extracting the foreground;
[0090] S4. Use the Canny operator to perform edge detection and set a threshold. Use the ellipse fitting algorithm to obtain the characteristic parameters of the citrus fruit. Convert back to the RGB color space to calculate the average color value. Combine the shape and color features to determine the position of the fruit neck. Calculate the pixel distance and convert it into actual length according to the calibration parameters.
[0091] S5. Arrange multiple sensors at the top and bottom of the citrus growing environment to collect temperature, humidity, and light intensity data in real time;
[0092] S6. Calculate the correlation coefficients between environmental parameters and fruit neck length and the correlation coefficients between image information and citrus fruit neck length, establish a linear model to analyze the significance between the variables, and screen out the key variables affecting fruit neck length based on the model.
[0093] Technical solution for lengthening the neck of citrus fruit:
[0094] Reasonable facility warming: Facility greenhouse covering can effectively increase the temperature during the flowering and young fruiting stages of citrus cultivation, but temperatures above 40 degrees will inhibit the growth of citrus plants, and temperatures exceeding 45 degrees for a long time will cause the entire citrus plant to wither and die. This cultivation model controls the ground temperature of citrus during the flowering and young fruiting stages between 35 degrees and 40 degrees in May and June by covering and removing the film, which can effectively lengthen the fruit neck without affecting the tree vigor.
[0095] Trunk cultivation mode: There is a temperature difference of nearly ten degrees between the top of the greenhouse and the ground. Normal "happy-type" citrus cultivation is mostly planted near the ground. The high-pole trunk cultivation mode is adopted, which can move the flowering and fruiting parts of the citrus to the top of the greenhouse, which can greatly increase the temperature of the citrus during the flowering and young fruit periods, and promote the elongation of the citrus fruit neck.
[0096] Proper temperature control delays flowering: The normal flowering period for greenhouses is mid-April, and the normal flowering period for open-air cultivation is late April or early May. This cultivation model uses a fully open film greenhouse management mode from February to April to reduce the effective accumulated temperature in the greenhouse and delay the flowering period of the ugly orange to early May.
[0097] Heavy pruning utilizes single, leafy flowers: Under normal facility cultivation, the clustered, leafless flowers of Ugly Mandarin oranges bloom early, generally around April 10th, while the single, leafy flowers bloom later, generally after April 25th. Stronger trees bloom later, while weaker trees bloom earlier. Based on this characteristic, cultivating single, leafy flowers using techniques such as "retraction" and "heavy pruning" can delay flowering by another 5-10 days by cultivating them as fruiting branches.
[0098] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.
Claims
1. An intelligent detection system for citrus fruit neck elongation, characterized in that: include: An image acquisition module is configured to obtain image information of the neck of the citrus fruit to be detected, and process the image information to obtain the neck length of the citrus fruit variety; An environmental parameter acquisition module is configured to acquire environmental parameters of the growth environment of the citrus to be tested; The data analysis module is electrically connected to the environmental parameter acquisition module and the image acquisition module respectively, and is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length.
2. The intelligent detection system for citrus fruit neck elongation according to claim 1, characterized in that: When the image acquisition module is configured to acquire image information of the citrus neck to be detected, it includes: The image acquisition module is also configured to obtain image information of the citrus fruit neck to be detected through a high-definition camera. An LED light board is set under each high-definition camera, and the LED light board is linked to the high-definition camera.
3. The intelligent detection system for citrus fruit neck elongation according to claim 2, characterized in that: When the image acquisition module processes the image information, it includes: The image acquisition module is further configured to perform denoising on the image information. The denoising adopts a median filtering algorithm, the filter window size is set to 3x3, and the contrast of the denoised image information is compared with a preset contrast of the image acquisition module, wherein: If the contrast is lower than a preset contrast of the image acquisition module, the image acquisition module improves the contrast of the image information after denoising by using histogram equalization; If the contrast is greater than or equal to a preset contrast of the image acquisition module, the image acquisition module does not improve the contrast of the image information after denoising by using histogram equalization.
4. The intelligent detection system for citrus fruit neck elongation according to claim 3, characterized in that: When the image acquisition module processes the image information, it includes: The image acquisition module is further configured to convert the image information from a default RGB color space to an HSV color space, and set an H channel threshold, an S channel threshold, and a V channel threshold, and generate a binary mask based on the H channel threshold, the S channel threshold, and the V channel threshold to obtain a binary image including the citrus neck area; The image acquisition module is further configured to calculate the gray level co-occurrence matrix of the binary image and extract the corresponding texture contrast and texture correlation, wherein: If the texture contrast of a certain area is lower than the preset texture contrast threshold, the area is judged as branch and leaf debris and removed from the foreground; If the texture correlation of a certain area is greater than or equal to a preset texture correlation threshold, the area is determined to be the surface of the citrus neck and is retained as the foreground.
5. The intelligent detection system for citrus fruit neck elongation according to claim 4, characterized in that: When the image acquisition module obtains the neck length of the citrus variety, it includes: The image acquisition module is further configured to use the Canny operator of the edge detection algorithm to extract the edge of the citrus fruit contour, set a threshold for edge detection, and use an ellipse fitting algorithm to fit the citrus fruit contour to obtain characteristic parameters of the citrus fruit; The image acquisition module is further configured to convert the image information into an RGB color space and calculate the average R, G, and B values of the citrus fruit region as a color feature vector; The image acquisition module is further configured to determine the position of the neck of the citrus fruit according to the shape characteristics and color characteristics of the citrus fruit, including: By analyzing the color changes and texture differences in the top area of the citrus fruit after the ellipse fitting, the starting point of the fruit neck is found. The position of the fruit stalk is searched along the citrus fruit outline to determine the end point of the fruit neck. The pixel distance of the fruit neck in the image is calculated and converted into actual length according to the calibration parameters of the camera.
6. The intelligent detection system for citrus fruit neck elongation according to claim 5, characterized in that: When the environmental parameter acquisition module is configured to acquire the environmental parameters of the growth environment of the citrus to be detected, it includes: The environmental parameter acquisition module is also configured to set multiple temperature sensors, humidity sensors and light intensity sensors at the top and bottom of the environment in which the citrus is located, and obtain temperature data, humidity data and light intensity data of the environment in which the citrus is located in real time.
7. The intelligent detection system for citrus fruit neck elongation according to claim 6, characterized in that: When the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes: The data analysis module is further configured to calculate the correlation coefficient between the environmental parameter and the length of the citrus neck, and the correlation coefficient between the image information and the length of the citrus neck.
8. The intelligent detection system for citrus fruit neck elongation according to claim 7, characterized in that: When the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes: The data analysis module is further configured to establish a linear model between the citrus neck length and the environmental parameters and image information, and analyze the significance of each variable.
9. The intelligent detection system for citrus fruit neck elongation according to claim 8, characterized in that: When the data analysis module is configured to obtain the correlation between the image information, the environmental parameters and the fruit neck length, it includes: The data analysis module is further configured to screen out key variables based on the feature importance of the linear model.
10. An intelligent detection method for citrus fruit neck elongation, characterized in that: An intelligent detection system for citrus fruit neck elongation as claimed in any one of claims 1 to 9, comprising the following steps: S1. Use high-definition cameras to acquire image information of the citrus fruit necks to be inspected. An LED light panel is installed below each camera to provide linked lighting. The image is denoised using a 3x3 median filter algorithm. The contrast is compared with a preset value to determine whether histogram equalization is performed to improve the contrast. S2, converting the processed image from RGB color space to HSV color space, setting H, S, and V channel thresholds to generate a binary mask to separate the area containing the citrus neck; S3. Calculate the gray-level co-occurrence matrix of the binary image, extract texture contrast and correlation, and determine whether the area is a branch or leaf fragment or the surface of a citrus fruit neck based on a preset threshold, thereby accurately extracting the foreground; S4. Use the Canny operator to perform edge detection and set a threshold. Use the ellipse fitting algorithm to obtain the characteristic parameters of the citrus fruit. Convert back to the RGB color space to calculate the average color value. Combine the shape and color features to determine the position of the fruit neck. Calculate the pixel distance and convert it into actual length according to the calibration parameters. S5. Arrange multiple sensors at the top and bottom of the citrus growing environment to collect temperature, humidity, and light intensity data in real time; S6. Calculate the correlation coefficients between environmental parameters and fruit neck length and the correlation coefficients between image information and citrus fruit neck length, establish a linear model to analyze the significance between the variables, and screen out the key variables affecting fruit neck length based on the model.