Flower sorting full-process supervision system based on big data

Abstract

The invention discloses a flower sorting full-process supervision system based on big data, which is characterized in that it comprises a supervision center, wherein the supervision center is communicatively connected with a data acquisition module, a data processing module, a feature extraction module, a flower classification module and a supervision module; the data acquisition module is used to divide the flower sorting full process to obtain a plurality of sorting sub-processes, and collect flower data in corresponding sorting sub-processes to obtain corresponding flower information; data processing is performed on the collected flower information to obtain corresponding standard images; feature extraction is performed based on the obtained standard images to obtain corresponding flower feature information; a classification model is constructed based on the obtained flower feature information, and a corresponding classification strategy is generated according to the constructed classification model; the corresponding classification process is supervised, and corresponding information feedback is taken according to the supervision result; the invention realizes efficient supervision and management of the flower sorting full process.

Description

Technical Field

[0001] The present invention relates to the field of big data, and in particular to a full-process supervision system for flower sorting based on big data. Background Art

[0002] With the upgrading of flower consumption, more and more flowers are entering households, and the flower industry has become a rapidly growing industry with huge potential.

[0003] Compared with existing technologies, the traditional flower sorting process has problems such as low efficiency, high error rate, and inability to monitor in real time, which leads to increased loss rate of flowers during transportation and affects the final sales quality. These are the problems we need to solve, and for this reason we provide a full-process flower sorting monitoring system based on big data. Summary of the invention

[0004] The purpose of the present invention is to provide a full-process supervision system for flower sorting based on big data.

[0005] The purpose of the present invention can be achieved by the following technical solutions: a flower sorting full-process supervision system based on big data, including a supervision center, wherein the supervision center is communicatively connected with a data acquisition module, a data processing module, a feature extraction module, a flower classification module and a supervision module;

[0006] The data collection module is used to divide the whole process of flower sorting, obtain several sorting sub-processes, and collect the flower data in the corresponding sorting sub-processes to obtain corresponding flower information;

[0007] The data processing module is used to process the collected flower information to obtain a corresponding standard image;

[0008] The feature extraction module is used to perform feature extraction based on the obtained standard image to obtain corresponding flower feature information, wherein the flower feature information includes first feature data and a flower feature map;

[0009] The flower classification module is used to construct a classification model based on the obtained flower feature information, and generate a corresponding classification strategy according to the constructed classification model;

[0010] The supervision module is used to supervise the corresponding classification process and take corresponding information feedback according to the supervision result.

[0011] Furthermore, the data collection module divides the whole process of flower sorting to obtain several sorting sub-processes, including:

[0012] The whole process of sorting corresponding flowers is collected, and the whole process of flower sorting shown includes harvesting, inspection, classification, cleaning, packaging and distribution; the obtained whole process of flower sorting is divided according to the actual flower production situation to obtain several sorting sub-processes.

[0013] Furthermore, the process of collecting the flower data in the corresponding sorting sub-process and obtaining the corresponding flower information includes:

[0014] A data acquisition unit is provided, wherein the data acquisition unit is provided with an acquisition node and a monitoring node. The acquisition node is used to collect data on the classification environment to which the corresponding sorting sub-process belongs to obtain corresponding environmental data; the monitoring node is used to collect images of flowers that need to be classified in the corresponding sorting sub-process to obtain corresponding flower images; the collected environmental data and flower images are summarized to obtain corresponding flower information, and the information is uploaded to the supervision center for storage.

[0015] Furthermore, the data processing module processes the collected flower information to obtain a corresponding standard image, which includes:

[0016] Read the collected flower image, perform image preprocessing on the obtained flower image, and after the image preprocessing is completed, mark the preprocessed flower image as a grayscale image;

[0017] Constructing a two-dimensional rectangular coordinate system, and mapping the obtained grayscale image into the constructed two-dimensional rectangular coordinate system;

[0018] Obtain the grayscale values ​​corresponding to all pixels in the corresponding grayscale image; obtain the maximum and minimum grayscale values ​​in each row of pixels in the corresponding grayscale image based on the obtained grayscale values; then, add the obtained maximum and minimum grayscale values ​​corresponding to each row of pixels to obtain the average value, and use the grayscale value corresponding to the obtained maximum grayscale value as the maximum grayscale level, and use the grayscale value corresponding to the obtained minimum grayscale value as the minimum grayscale level;

[0019] Then, pixel scaling is performed on the corresponding grayscale image based on the obtained maximum grayscale level and minimum grayscale level, and the corresponding scaled grayscale image is marked as a standard image.

[0020] Furthermore, the process of performing feature extraction based on the obtained standard image to obtain corresponding first feature data includes:

[0021] Read the obtained standard image, obtain all grayscale values ​​in the corresponding standard image, and construct a corresponding grayscale histogram based on the grayscale values;

[0022] Obtaining the gray level corresponding to the corresponding standard image based on the obtained gray histogram;

[0023] An initial threshold value T0 is set, the grayscale value of the pixel point in the corresponding standard image is compared with the corresponding initial threshold value, and the entire image area of ​​the corresponding standard image is divided into a corresponding first area and a second area according to the comparison result, and the pixel proportions corresponding to the corresponding first area and the second area are respectively obtained, and the variance coefficient corresponding to the corresponding standard image is obtained according to the obtained pixel proportions;

[0024] A gray value is selected based on the gray value range in the obtained gray histogram, and is used as a new initial threshold to obtain a variance coefficient corresponding to the corresponding gray level;

[0025] Count the variance coefficients corresponding to all gray levels, and use the gray value corresponding to the maximum variance coefficient as the final gray threshold;

[0026] Furthermore, the corresponding standard image is converted into a binary image based on the obtained final grayscale threshold, and features are extracted from the obtained binary image based on an image segmentation algorithm to obtain corresponding first feature data.

[0027] Furthermore, the process of performing feature extraction based on the obtained standard image to obtain the corresponding flower feature image includes:

[0028] The corresponding standard image is gray-scaled and converted to obtain a corresponding color image; the RGB color space is extracted based on OPENCV software to obtain a color channel image corresponding to the corresponding color image, wherein the color channel image includes an R channel image, a G channel image and a B channel image;

[0029] Constructing a corresponding color feature matrix based on the obtained color channel image, wherein the color feature matrix includes a first-order matrix and a second-order matrix;

[0030] The obtained first-order matrix μ1 and second-order matrix μ2 are input into the pre-trained ResNet-18 model to obtain the corresponding R channel feature image;

[0031] The obtained G channel image and B channel image are processed respectively by the same method to obtain the corresponding G channel characteristic image and B channel characteristic image;

[0032] The obtained R channel characteristic image, G channel characteristic image and B channel characteristic image are superimposed to obtain a corresponding flower characteristic image.

[0033] Furthermore, the flower classification module constructs a classification model based on the obtained flower feature information, and generates a corresponding classification strategy according to the constructed classification model, including:

[0034] Constructing a classification model, inputting the obtained flower feature information into the constructed classification model, and obtaining corresponding output results;

[0035] The obtained output results are fed back to the supervision center, which generates a corresponding classification control strategy based on the received output results, controls the robot arm corresponding to the corresponding sorting conveyor belt to classify flowers based on the obtained classification control strategy, and supervises the corresponding classification process.

[0036] Furthermore, the process of supervising the corresponding classification process of the supervision module and taking corresponding information feedback according to the supervision result includes:

[0037] Set a monitoring period and read the classification strategy obtained in the corresponding monitoring period; obtain all classification categories in the corresponding monitoring period and the estimated number of flowers corresponding to the corresponding categories according to the obtained classification strategy, and obtain the actual number of flowers corresponding to each classification category in the corresponding monitoring period;

[0038] Obtain output results corresponding to corresponding classification categories, classify flowers of corresponding classification categories into qualified products and unqualified products according to the output results, and obtain the qualified rate of flowers in the corresponding monitoring period according to the actual number of flowers corresponding to the corresponding qualified products and unqualified products;

[0039] Set a qualified threshold, compare the corresponding qualified rate with the qualified threshold, and take corresponding measures based on the comparison results;

[0040] If the qualified rate is lower than the qualified threshold, the environmental data in the corresponding flower data is read, compared with the pre-set environmental threshold, and corresponding measures are taken according to the comparison results;

[0041] If the qualified rate is not lower than the qualified threshold, the corresponding ratio of the estimated number of flowers to the actual number of flowers will be obtained and marked as the classification coefficient;

[0042] Set the classification threshold range, compare the obtained classification coefficient with the classification threshold range, and take corresponding measures based on the comparison results.

[0043] Compared with the prior art, the beneficial effects of the present invention are: by integrating functional modules such as data collection, processing, feature extraction and supervision, efficient supervision and management of the whole process of flower sorting is realized; it not only improves the accuracy and efficiency of flower sorting, but also provides strong support for flower production and sales. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of the present invention. DETAILED DESCRIPTION

[0045] like Figure 1 As shown, a flower sorting full-process supervision system based on big data includes a supervision center, and the supervision center is communicatively connected with a data acquisition module, a data processing module, a feature extraction module, a flower classification module and a supervision module;

[0046] The data collection module is used to divide the whole process of flower sorting, obtain several sorting sub-processes, and collect the flower data in the corresponding sorting sub-processes to obtain corresponding flower information;

[0047] It should be further explained that, in the specific implementation process, the data acquisition module divides the whole process of flower sorting, and the process of obtaining several sorting sub-processes includes:

[0048] Taking any variety of flowers as an example, the whole process of sorting the corresponding flowers is collected, and the whole process of flower sorting shown includes harvesting, inspection, classification, cleaning, packaging and distribution; the obtained whole process of flower sorting is divided according to the actual flower production situation to obtain several sorting sub-processes;

[0049] It should be further explained that, in the specific implementation process, the process of collecting the flower data in the corresponding sorting sub-process and obtaining the corresponding flower information includes:

[0050] A data acquisition unit is provided, wherein the data acquisition unit is provided with an acquisition node and a monitoring node, and the acquisition node and the monitoring node are respectively provided in the sorting transmission of the corresponding sub-process to the periphery; the acquisition node is used to collect data on the classification environment to which the corresponding sorting sub-process belongs, and obtain the corresponding ambient temperature, ambient humidity, ambient light intensity, etc.; the monitoring node is used to collect images of flowers that need to be classified in the corresponding sorting sub-process, and obtain corresponding flower images; the collected environmental data and flower images are summarized to obtain corresponding flower information, and the information is uploaded to the supervision center for storage.

[0051] The data processing module is used to process the collected flower information to obtain a corresponding standard image;

[0052] It should be further explained that, in the specific implementation process, the data processing module processes the collected flower information to obtain the corresponding standard image, which includes:

[0053] Reading the collected flower images, and performing image preprocessing on the obtained flower images, wherein the image preprocessing includes image denoising, image enhancement, and image grayscale conversion;

[0054] The image denoising refers to reducing or eliminating the noise in the collected flower image based on Gaussian filtering technology; the image enhancement technology refers to enhancing the collected flower image by increasing the contrast or resolution; the image hole filling refers to filling the holes caused by factors such as light during the shooting process; the image graying refers to graying the collected flower image;

[0055] After the image preprocessing is completed, the preprocessed flower image is marked as a grayscale image;

[0056] Construct a two-dimensional rectangular coordinate system, and map the obtained grayscale image into the constructed two-dimensional rectangular coordinate system, then the coordinates of the pixel points in the corresponding grayscale image can be marked as (x, y);

[0057] Obtain the grayscale values ​​corresponding to all pixels in the corresponding grayscale image, and record them as G(x, y); obtain the maximum and minimum grayscale values ​​in each row of pixels in the corresponding grayscale image based on the obtained grayscale values; then, add the maximum and minimum grayscale values ​​corresponding to each row of pixels to obtain the average value, and take the grayscale value corresponding to the maximum grayscale value as the maximum grayscale level, which is recorded as Gmax; take the grayscale value corresponding to the minimum grayscale value as the minimum grayscale level, which is recorded as Gmin;

[0058] Then, pixel scaling is performed on the corresponding grayscale image based on the obtained maximum grayscale level and minimum grayscale level; the corresponding mathematical formula is as follows:

[0059]

[0060] Where H(x, y) represents the gray value of the pixel (x, y) after scaling;

[0061] The corresponding scaled grayscale image is marked as the canonical image.

[0062] The feature extraction module is used to perform feature extraction based on the obtained standard image to obtain corresponding flower feature information, wherein the flower feature information includes first feature data and a flower feature map;

[0063] It should be further explained that, in the specific implementation process, the feature extraction module performs feature extraction based on the obtained standard image, and the process of obtaining the corresponding flower feature information includes:

[0064] Read the obtained standard image, obtain all grayscale values ​​in the corresponding standard image, and construct a corresponding grayscale histogram based on the grayscale values;

[0065] Based on the obtained grayscale histogram, the grayscale level corresponding to the corresponding standard image is obtained, and it is numbered and recorded as i, where i = 0, 1, 2, ..., n, n, ≥ 0 and n is an integer; the number of pixels corresponding to the corresponding grayscale level i is recorded as Ni;

[0066] An initial threshold value T0 is set, the grayscale value of the pixel in the corresponding standard image is compared with the corresponding initial threshold value, and the entire image area of ​​the corresponding standard image is divided into a corresponding first area and a second area according to the comparison result, and the pixel proportions corresponding to the corresponding first area and the second area are respectively obtained, and the obtained pixel proportions are recorded as P0 and P1 respectively; wherein the pixel proportion refers to the ratio of the total number of pixel points in the corresponding first area to the total number of pixel points in the corresponding standard image, and P0+P1=1;

[0067] Based on the obtained pixel ratio, the variance coefficient corresponding to the corresponding standard image is obtained and recorded as D;

[0068] Where D = P0 × (E0-E) 2 +P1×(E1-E) 2 ;

[0069] Wherein, E0, E1, and E represent the grayscale average values ​​corresponding to the pixels in the first region, the second region, and the standard image, respectively;

[0070] Then, a gray value is selected based on the gray value range in the obtained gray histogram, and is used as a new initial threshold to obtain a variance coefficient corresponding to the corresponding gray level;

[0071] Count the variance coefficients corresponding to all gray levels, and use the gray value corresponding to the maximum variance coefficient as the final gray threshold;

[0072] Then, based on the obtained final grayscale threshold, the corresponding standard image is converted into a binary image, and based on the image segmentation algorithm, feature extraction is performed on the obtained binary image to obtain corresponding first feature data, wherein the first feature data includes but is not limited to contour features, area features, and shape features;

[0073] The corresponding standard image is gray-scaled and converted to obtain a corresponding color image; the RGB color space is extracted based on OPENCV software to obtain a color channel image corresponding to the corresponding color image, wherein the color channel image includes an R channel image, a G channel image and a B channel image;

[0074] Constructing a corresponding color feature matrix based on the obtained color channel image, wherein the color feature matrix includes a first-order matrix and a second-order matrix;

[0075] Taking the R channel image as an example, the first-order matrix and the second-order matrix corresponding to the corresponding R channel image are marked as μ1 and μ2 respectively;

[0076] in,

[0077]

[0078] Where R(i, j) represents the pixel value corresponding to the pixel point coordinate (i, j) in the corresponding R channel image;

[0079] The obtained first-order matrix μ1 and second-order matrix μ2 are input into the pre-trained ResNet-18 model to obtain the corresponding R channel feature image;

[0080] The obtained G channel image and B channel image are processed respectively by the same method to obtain the corresponding G channel characteristic image and B channel characteristic image;

[0081] The obtained R channel characteristic image, G channel characteristic image and B channel characteristic image are superimposed to obtain a corresponding flower characteristic image.

[0082] The flower classification module is used to construct a classification model based on the obtained flower feature information, and generate a corresponding classification strategy according to the constructed classification model;

[0083] It should be further explained that, in the specific implementation process, the flower classification module constructs a classification model based on the obtained flower feature information, and generates a corresponding classification strategy according to the constructed classification model, including:

[0084] Based on big data technology, historical flower feature data corresponding to several categories of flowers are obtained, and corresponding training sets and test sets are constructed according to the obtained historical flower feature data;

[0085] An initial classification model is built based on neural network technology, and the corresponding initial classification model is iteratively trained based on the obtained training set until the loss function is stable, and the model parameters are saved; the initial classification model completed by iterative training is verified based on the obtained test set, and the corresponding classification model is obtained after the verification is passed;

[0086] Input the obtained flower characteristic information into the constructed classification model to obtain corresponding output results, wherein the output results are the classification category, quality, water content and other related attributes of the corresponding flowers;

[0087] The obtained output results are fed back to the supervision center, and the supervision center generates a corresponding classification control strategy based on the received output results, and controls the robot arm corresponding to the corresponding sorting conveyor belt to classify flowers according to the obtained classification control strategy, and supervises the corresponding classification process.

[0088] The supervision module is used to supervise the corresponding classification process and take corresponding information feedback according to the supervision results;

[0089] It should be further explained that, in the specific implementation process, the supervision module supervises the corresponding classification process, and takes corresponding information feedback according to the supervision results, including:

[0090] Set the monitoring period and read the classification strategy obtained in the corresponding monitoring period; obtain all the classification categories and the estimated number of flowers in the corresponding categories in the corresponding monitoring period according to the obtained classification strategy, and mark them as L and X respectively 分 ;

[0091] After the corresponding monitoring period ends, obtain the actual number of flowers corresponding to each classification category in the corresponding monitoring period and mark it as X 实 ;

[0092] Taking a certain category as an example, the output result corresponding to the corresponding classification category is obtained, and the flowers of the corresponding classification category are divided into qualified products and unqualified products according to the output result, and the qualified rate of flowers in the corresponding monitoring period is obtained according to the actual number of flowers corresponding to the corresponding qualified products and unqualified products;

[0093] Set a qualified threshold, compare the corresponding qualified rate with the qualified threshold, and take corresponding measures based on the comparison results;

[0094] If the qualified rate is lower than the qualified threshold, the supervision center reads the environmental data in the corresponding flower data and compares it with the preset environmental threshold. If the corresponding environmental data does not meet the environmental threshold, the environmental abnormality feedback is sent to the corresponding supervision center to remind the management personnel of the supervision center to adjust the environmental data of the corresponding classification sub-process. It should be further explained that in the specific implementation process, the environmental threshold is set in advance by the staff based on work experience, and different sorting sub-processes may have partial differences. If the corresponding environmental data meets the environmental threshold, a corresponding process warning is generated and fed back to the corresponding management personnel to remind the management personnel to check the corresponding flower sorting process.

[0095] If the pass rate is not lower than the pass threshold, the corresponding estimated number of flowers X will be obtained 分 and the actual number of flowers X 实 and label it as the classification coefficient;

[0096] Set a classification threshold range and compare the obtained classification coefficient with the classification threshold range;

[0097] If the classification coefficient is higher than the maximum value of the classification threshold range, no other operations are performed;

[0098] If the classification threshold is within the classification threshold range, the corresponding classification category will be marked and fed back to the supervision center, and the supervision center will send management personnel to manually screen the sorting and transmission process belonging to the corresponding classification category;

[0099] If the classification threshold is lower than the minimum value of the classification threshold range, a corresponding misclassification warning is generated and fed back to the supervision center, and the transmission speed of the corresponding sorting conveyor belt is reduced.

[0100] It should be further explained that, in the specific implementation process, the present invention defaults to taking the classification sub-process within the corresponding sorting full process as an example.

[0101] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A full-process supervision system for flower sorting based on big data, characterized by: It includes a supervision center, which is communicatively connected with a data acquisition module, a data processing module, a feature extraction module, a flower classification module and a supervision module; The data collection module is used to divide the whole process of flower sorting, obtain several sorting sub-processes, and collect the flower data in the corresponding sorting sub-processes to obtain corresponding flower information; The data processing module is used to process the collected flower information, and the process of obtaining the corresponding standard image includes: Read the collected flower image, perform image preprocessing on the obtained flower image, and after the image preprocessing is completed, mark the preprocessed flower image as a grayscale image; Constructing a two-dimensional rectangular coordinate system, and mapping the obtained grayscale image into the constructed two-dimensional rectangular coordinate system; Obtain the grayscale values ​​corresponding to all pixels in the corresponding grayscale image; obtain the maximum and minimum grayscale values ​​in each row of pixels in the corresponding grayscale image based on the obtained grayscale values; then, add the obtained maximum and minimum grayscale values ​​corresponding to each row of pixels to obtain the average value, and use the grayscale value corresponding to the obtained maximum grayscale value as the maximum grayscale level, and use the grayscale value corresponding to the obtained minimum grayscale value as the minimum grayscale level; Then, pixel scaling is performed on the corresponding grayscale image based on the obtained maximum grayscale level and minimum grayscale level, and the corresponding scaled grayscale image is marked as a standard image; The feature extraction module is used to perform feature extraction based on the obtained standard image to obtain corresponding flower feature information, wherein the flower feature information includes first feature data and a flower feature map; The process of obtaining the corresponding first characteristic data includes: Read the obtained standard image, obtain all grayscale values ​​in the corresponding standard image, and construct a corresponding grayscale histogram based on the grayscale values; Obtaining the gray level corresponding to the corresponding standard image based on the obtained gray histogram; An initial threshold value T0 is set, and the grayscale value of the pixel point in the corresponding standard image is compared with the corresponding initial threshold value, and the entire image area of ​​the corresponding standard image is divided into a corresponding first area and a second area according to the comparison result, and the pixel proportions corresponding to the corresponding first area and the second area are respectively obtained, and the obtained pixel proportions are respectively recorded as P0 and P1; the pixel proportions refer to the ratios of the total number of pixel points in the corresponding area to the total number of pixel points in the corresponding standard image, and P0+P1=1; the variance coefficient corresponding to the corresponding standard image is obtained according to the obtained pixel proportions, and is recorded as D; in, ; Wherein, E0, E1, and E represent the grayscale average values ​​corresponding to the pixels in the first region, the second region, and the standard image, respectively; A gray value is selected based on the gray value range in the obtained gray histogram, and is used as a new initial threshold to obtain a variance coefficient corresponding to the corresponding gray level; Count the variance coefficients corresponding to all gray levels, and use the gray value corresponding to the maximum variance coefficient as the final gray threshold; Then, based on the obtained final grayscale threshold, the corresponding standard image is converted into a binary image, and based on the image segmentation algorithm, feature extraction is performed on the obtained binary image to obtain corresponding first feature data; The process of obtaining the corresponding flower feature image includes: The corresponding standard image is gray-scaled and converted to obtain a corresponding color image; the RGB color space is extracted based on OPENCV software to obtain a color channel image corresponding to the corresponding color image, wherein the color channel image includes an R channel image, a G channel image and a B channel image; Constructing a corresponding color feature matrix based on the obtained color channel image, wherein the color feature matrix includes a first-order matrix and a second-order matrix; The obtained first-order matrix and second-order matrix are input into the pre-trained ResNet-18 model to obtain the corresponding R channel feature image; The obtained G channel image and B channel image are processed respectively by the same method to obtain the corresponding G channel characteristic image and B channel characteristic image; The obtained R channel characteristic image, G channel characteristic image and B channel characteristic image are superimposed to obtain a corresponding flower characteristic image; The flower classification module is used to construct a classification model based on the obtained flower feature information, and generate a corresponding classification strategy according to the constructed classification model; The supervision module is used to supervise the corresponding classification process and take corresponding information feedback according to the supervision result.

2. According to the big data-based flower sorting full-process supervision system of claim 1, it is characterized in that: The data acquisition module divides the whole process of flower sorting to obtain several sorting sub-processes, including: The whole process of sorting corresponding flowers is collected, and the whole process of flower sorting shown includes harvesting, inspection, classification, cleaning, packaging and distribution; the obtained whole process of flower sorting is divided according to the actual flower production situation to obtain several sorting sub-processes.

3. The flower sorting full-process supervision system based on big data according to claim 2 is characterized in that: The process of collecting the flower data in the corresponding sorting sub-process and obtaining the corresponding flower information includes: A data acquisition unit is provided, wherein the data acquisition unit is provided with an acquisition node and a monitoring node. The acquisition node is used to collect data on the classification environment to which the corresponding sorting sub-process belongs, obtain corresponding environmental data, and obtain corresponding environmental temperature, environmental humidity and environmental light intensity; the monitoring node is used to collect images of flowers that need to be classified in the corresponding sorting sub-process, and obtain corresponding flower images; the collected environmental data and flower images are summarized to obtain corresponding flower information, and the information is uploaded to the supervision center for storage.

4. The flower sorting full-process supervision system based on big data according to claim 3 is characterized in that: The flower classification module constructs a classification model based on the obtained flower feature information, and generates a corresponding classification strategy according to the constructed classification model, including: Constructing a classification model, inputting the obtained flower feature information into the constructed classification model, and obtaining corresponding output results; The obtained output results are fed back to the supervision center, and the supervision center generates a corresponding classification control strategy based on the received output results, and controls the robot arm corresponding to the corresponding sorting conveyor belt to classify flowers according to the obtained classification control strategy, and supervises the corresponding classification process.

5. The flower sorting full-process supervision system based on big data according to claim 4 is characterized in that: The supervision module supervises the classification process and takes corresponding information feedback according to the supervision result, including: Set a monitoring period and read the classification strategy obtained in the corresponding monitoring period; obtain all classification categories in the corresponding monitoring period and the estimated number of flowers corresponding to the corresponding categories according to the obtained classification strategy, and obtain the actual number of flowers corresponding to each classification category in the corresponding monitoring period; Obtain output results corresponding to corresponding classification categories, classify flowers of corresponding classification categories into qualified products and unqualified products according to the output results, and obtain the qualified rate of flowers in the corresponding monitoring period according to the actual number of flowers corresponding to the corresponding qualified products and unqualified products; Set a qualified threshold, compare the corresponding qualified rate with the qualified threshold, and take corresponding measures based on the comparison results; If the qualified rate is lower than the qualified threshold, the environmental data in the corresponding flower data is read, compared with the pre-set environmental threshold, and corresponding measures are taken according to the comparison results; If the qualified rate is not lower than the qualified threshold, the corresponding ratio of the estimated number of flowers to the actual number of flowers will be obtained and marked as the classification coefficient; Set the classification threshold range, compare the obtained classification coefficient with the classification threshold range, and take corresponding measures based on the comparison results.

Images