Green plant health assessment method and device in confined space

Abstract

The present application discloses a method and device for evaluating the health of green plants in a confined space, and relates to the field of electronic digital data processing. The method utilizes the characteristic that the withering, falling off, wilting and death of plant leaves, branches, flowers and fruits are all shrinkage processes caused by water reduction. Its intuitive visual effect is the reduction of the visual area of ​​the plant under multiple viewing angles. The plant can show this characteristic when it is initially short of water. As the degree of water shortage increases, the visual area will continue to decrease, that is, the degree of water shortage and the visual area are positively correlated, so that the present application can capture this characteristic through at least two viewing angles, and there is no need to identify the plant variety, and the health status can be judged only by the visual area.

Description

Technical Field

[0001] The present application relates to the technical field of electronic digital data processing, and in particular to a method and device for evaluating the health of green plants in a confined space. Background Art

[0002] In modern large-scale buildings, green plants have begun to be planted in restricted spaces (such as glass covers, partitions, curtain walls, etc.) to improve the indoor environment, provide fresh air to the indoor environment, and increase green and ecological sensory elements. Indoor cultivation of different plants can be achieved by adjusting the cultivation temperature, humidity, wind speed, etc. Planting green plants in restricted spaces is currently the mainstream greening and oxygenation method for indoor public environments and curtain walls.

[0003] Currently, the maintenance of the eco-capsule is through regular and unified artificial irrigation to cultivate the plants inside. Due to the different habits of plants and the fact that most staff of large buildings have not formulated corresponding irrigation plans for different varieties of plants, the plants are prone to negative conditions such as yellowing, withering, and falling of leaves, branches, flowers, and fruits.

[0004] Generally, when green plants show the above symptoms in the early stage, human intervention is required to treat the plants. However, the current evaluation of the health status of green plants is usually through manual visual observation. Intervention will only be carried out when obvious visual phenomena such as large-scale withering and large-scale shedding of green plants are observed. At this time, the green plants are often in an irreversible withering process, which results in the staff having to replace the entire green plant, and the health of the green plants cannot be guaranteed. Summary of the invention

[0005] The main purpose of the present application is to provide a method and device for evaluating the health of green plants in a confined space, so as to solve the problem in the prior art that the green plants in the confined space are not properly maintained, resulting in the health status and life span of the green plants not being guaranteed.

[0006] To achieve the above objectives, this application provides the following technical solutions:

[0007] A method for evaluating the health of green plants in a confined space, wherein the confined space is enclosed by four transparent baffles covered by a cover plate and a matrix in sequence from top to bottom, and green plants are planted on the matrix. The method for evaluating the health of green plants comprises:

[0008] Step S1, acquiring a plurality of visual images of the green plant from at least two viewing angles at intervals of a preset time period;

[0009] Step S2, obtaining all plant organs in each visual image respectively through a target detection algorithm;

[0010] Step S3, classifying all plant organs by a classification algorithm, and obtaining an organ number based on each preset plant organ type;

[0011] Step S4, respectively obtaining the image area value of each plant organ, and obtaining the sum of all the image area values;

[0012] Step S5, defining the number of organs in all preset time periods based on the same preset plant organ type as an independent variable, defining the sum value as a dependent variable, and defining the linear regression relationship between all independent variables and the dependent variable based on the same preset time period through a multiple linear regression model;

[0013] Step S6, calculating the regression coefficient of the linear regression relationship by the least square method and substituting it into the multivariate linear regression model to obtain an evaluation model;

[0014] Step S7, obtaining a real-time visual image of the green plant, and obtaining all real-time plant organs in the real-time visual image through the target detection algorithm;

[0015] Step S8, substituting all real-time plant organs into the evaluation model to obtain a real-time sum value;

[0016] Step S9, determining whether the real-time sum value is greater than the sum value, and if so, determining that the green plant is in a healthy state.

[0017] As a further improvement of the present application, step S2, obtaining all plant organs in each visual image respectively through a target detection algorithm, includes:

[0018] Step S21, dividing the current visual image into a plurality of grids on average;

[0019] Step S22, predicting a number of bounding boxes for all plant organs based on all grids according to the target detection algorithm;

[0020] Step S23, respectively obtaining the confidence of each bounding box, obtaining the bounding box with the largest confidence and marking it as a first-order bounding box;

[0021] Step S24, calculating the intersection-over-union ratio of each other bounding box with the first-order bounding box;

[0022] Step S25, selecting a bounding box whose intersection-over-union ratio is greater than or equal to a preset threshold as a second-order bounding box;

[0023] Step S26, obtaining a second-order bounding box with the highest confidence and defining it as the plant organ.

[0024] As a further improvement of the present application, in step S3, all plant organs are classified by a classification algorithm, and an organ number is obtained based on each preset plant organ type, including:

[0025] Step S31: Pack all plant organs in the same preset time period into a data set to be classified , For the A data set to be classified for a preset time period, For the The first plant organs, is the number of all plant organs;

[0026] Step S32, defining a plant organ category set according to the number of the preset plant organ types ,in, For the category set The preset plant organ types, is the number of the preset plant organ types;

[0027] Step S33, calculate the first The conditional probability of each plant organ in the dataset to be classified under each preset plant organ type:

[0028] (1);

[0029] in, For the Preset plant organ type The first The conditional probability of a plant organ; For the The marginal probability of a preset plant organ type; For the Under the preset plant organ type, The first The conditional probability of a plant organ;

[0030] Step S34: All plant organs in the dataset to be classified for a preset time period are classified into the preset plant organ type with the highest conditional probability, and the first A classification table of plant organs for a preset time period;

[0031] Step S35, counting the number of organs under each preset plant organ type.

[0032] As a further improvement of the present application, step S5, based on the same preset plant organ type, the number of organs in all preset time periods is defined as an independent variable, the sum value is defined as a dependent variable, and the linear regression relationship between all independent variables and the dependent variable is defined based on the same preset time period through a multiple linear regression model, including:

[0033] Step S51, defining the linear regression relationship between the dependent variable and all independent variables according to formula (2):

[0034] (2);

[0035] in, For the A dependent variable with a preset time period; is the intercept of the linear regression relationship; For the The first of the preset time periods Linear regression coefficients for each plant organ; is the random error of the linear regression relationship.

[0036] As a further improvement of the present application, step S6, calculating the regression coefficient of the linear regression relationship by the least squares method and substituting it into the multivariate linear regression model to obtain an evaluation model, including:

[0037] Step S61, calculate the first The regression coefficient of the linear regression relationship for a preset time period is:

[0038] (3);

[0039] in, for An estimated value of for The transposed matrix of .

[0040] As a further improvement of the present application, step S9, judging whether the real-time sum value is greater than the sum value, then includes:

[0041] Step S10, if not, determining that the green plant is in an unhealthy state;

[0042] Step S20, respectively acquiring color data of the plant organ in each preset time period by using the target detection algorithm;

[0043] Step S30, obtaining the current average value of all color data of the same preset plant organ type in the current preset time period;

[0044] Step S40, obtaining the historical average value of all color data of the same preset plant organ type in the historical time period based on the preset backoff step number;

[0045] Step S50, determining the difference between the current average value and the historical average value;

[0046] Step S60: marking the preset plant organ type whose difference is greater than or equal to the preset difference threshold as an unhealthy organ type.

[0047] As a further improvement of the present application, in step S60, the preset plant organ type whose difference is greater than or equal to the preset difference threshold is marked as an unhealthy organ type, and then, the following steps are included:

[0048] Step S100, sending the unhealthy organ type to an external receiving end.

[0049] To achieve the above objectives, this application also provides the following technical solutions:

[0050] A green plant health assessment device in a confined space, the green plant health assessment device is applied to the green plant health assessment method as described above, and the green plant health assessment device comprises:

[0051] A green plant visual image acquisition module, used to acquire a plurality of visual images of the green plant from at least two viewing angles at intervals of a preset time period;

[0052] A plant organ acquisition module is used to acquire all plant organs in each visual image through a target detection algorithm;

[0053] A plant organ classification module, used to classify all plant organs through a classification algorithm, and obtain an organ number based on each preset plant organ type;

[0054] An image area value and sum value acquisition module is used to respectively acquire the image area value of each plant organ and acquire the sum value of all image area values;

[0055] A linear regression relationship definition module, used to define the number of organs in all preset time periods based on the same preset plant organ type as an independent variable, define the sum value as a dependent variable, and define the linear regression relationship between all independent variables and the dependent variable based on the same preset time period through a multivariate linear regression model;

[0056] An evaluation model acquisition module, used to calculate the regression coefficient of the linear regression relationship by the least square method and substitute it into the multivariate linear regression model to obtain an evaluation model;

[0057] A real-time plant organ acquisition module, used to acquire the real-time visual image of the green plant, and acquire all real-time plant organs of the real-time visual image through the target detection algorithm;

[0058] A real-time sum value acquisition module is used to substitute all real-time plant organs into the evaluation model to obtain a real-time sum value;

[0059] The real-time sum value judgment module is used to judge whether the real-time sum value is greater than the sum value, and if so, it is judged that the green plant is in a healthy state.

[0060] To achieve the above objectives, this application also provides the following technical solutions:

[0061] An electronic device includes a processor and a memory coupled to the processor, wherein the memory stores program instructions executable by the processor; when the processor executes the program instructions stored in the memory, the green plant health assessment method as described above is implemented.

[0062] To achieve the above objectives, this application also provides the following technical solutions:

[0063] A storage medium stores program instructions, which, when executed by a processor, can implement the green plant health assessment method described above.

[0064] The present application obtains several visual images of green plants from at least two perspectives at intervals of preset time periods; obtains all plant organs in each visual image respectively through a target detection algorithm; classifies all plant organs through a classification algorithm, and obtains an organ number based on each preset plant organ type; obtains the image area value of each plant organ respectively, and obtains the sum of all image area values; defines the organ number of all preset time periods based on the same preset plant organ type as an independent variable, and defines the sum value as a dependent variable; defines the linear regression relationship between all independent variables and dependent variables based on the same preset time period through a multivariate linear regression model; calculates the regression coefficient of the linear regression relationship through the least squares method and substitutes it into the multivariate linear regression model to obtain an evaluation model; obtains real-time visual images of green plants, and obtains all real-time plant organs in the real-time visual images through a target detection algorithm; substitutes all real-time plant organs into the evaluation model to obtain a real-time sum value; determines whether the real-time sum value is greater than the sum value, and if so, determines that the green plants are in a healthy state. The present application utilizes the characteristic that the withering, falling off, wilting and death of plant leaves, branches, flowers and fruits are all shrinkage processes caused by water reduction. The intuitive visual effect is the reduction of the visual area of ​​the plant under multiple viewing angles. The plant can show this characteristic when it is initially short of water. As the degree of water shortage increases, the visual area will continue to decrease, that is, the degree of water shortage and the visual area are positively correlated. This allows the present application to capture this characteristic through at least two viewing angles, and there is no need to identify the plant variety. The health status can be judged only by the visual area. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a schematic diagram of the steps of an embodiment of a method for evaluating the health of green plants in a confined space of the present application;

[0066] Figure 2 This is a functional module diagram of an embodiment of a green plant health assessment device in a confined space of the present application;

[0067] Figure 3 This is a schematic diagram of the structure of an embodiment of the electronic device of the present application;

[0068] Figure 4 This is a schematic diagram of the structure of an embodiment of the storage medium of the present application. DETAILED DESCRIPTION

[0069] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0070] The terms "first", "second" and "third" in this application are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined as "first", "second" and "third" can explicitly or implicitly include at least one of the features. In the description of this application, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. All directional indications (such as up, down, left, right, front, back...) in the embodiments of this application are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.

[0071] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0072] like Figure 1 As shown, this embodiment provides an embodiment of a method for evaluating the health of green plants in a confined space. In this embodiment, the confined space is surrounded by four transparent baffles that are covered by a cover plate and a matrix in sequence from top to bottom, and green plants are planted on the matrix.

[0073] Preferably, this embodiment can be applied to an ecological cabin, which includes a hollow transparent column covered by an air circulation component and an oxygenation component in sequence from top to bottom, a substrate laid on the oxygenation component, and healthy plants located in the hollow transparent column and planted on the substrate.

[0074] Preferably, the eco-capsule is a prior art, and the eco-capsule is fixed between a first structural layer and a second structural layer at different heights, such as a floor and a ceiling, and the eco-capsule comprises:

[0075] The oxygen-enriching component includes a planting trough fixedly mounted on the floor, healthy plants planted on the planting trough, and a transparent enclosure fixedly mounted on the edge of the planting trough, wherein the transparent enclosure encloses the above-mentioned hollow transparent column.

[0076] The air circulation component includes a shell covering the top end of a hollow transparent column and a first air circulation component fixedly installed in the shell. The shell is fixedly connected to the transparent enclosure to form an air duct in the transparent enclosure. The first air inlet end of the first air circulation component passes through the shell and is connected to the air duct. The first air outlet end of the first air circulation component passes through the shell and is connected to the indoor space. The first air circulation component guides external air to pass through the air inlet, the air duct, the first air inlet end, and the first air outlet end in sequence to guide external air and oxygen produced by plants to the indoor space.

[0077] Preferably, the air circulation component also includes a second air circulation component fixedly installed in the shell, the second air inlet end of the second air circulation component passes through the shell and is connected to the indoor space, the second air outlet end of the second air circulation component passes through the shell and is connected to the outdoors, and the second air circulation component guides the air in the indoor space through the second air inlet end and the second air outlet end in sequence to guide the air in the indoor space to the outdoors.

[0078] Preferably, the oxygen enrichment component also includes a drip irrigation pipe fixedly installed in the planting trough, one end of the drip irrigation pipe is connected to an external water source, and the other end of the drip irrigation pipe is closed. At least one water outlet is opened on the side wall of the drip irrigation pipe, and a solenoid valve for conducting or blocking the external water source is fixedly installed in the drip irrigation pipe. The solenoid valve is located between the water outlet and the external water source, and the solenoid valve is opened or closed according to external instructions to conduct the external water source to the water outlet or block the external water source.

[0079] It is worth noting that the structure of the above-mentioned eco-cabin is already existing technology. For details, please refer to the patent "An air circulation device based on curtain wall", application number "202322289519.3". This embodiment does not improve the structure of the eco-cabin, but is a conventional application of the eco-cabin structure.

[0080] Specifically, the green plant health evaluation method includes the following steps:

[0081] Step S1, acquiring a plurality of visual images of green plants from at least two viewing angles at intervals of a preset time period.

[0082] Preferably, the preset time period can be set to one natural day, two natural days, etc., but should not be too long.

[0083] Step S2, obtaining all plant organs in each visual image respectively through a target detection algorithm.

[0084] Preferably, the target detection algorithm can be implemented by target detection algorithms VJ, HOG, DPMDetector; deep learning two-stage target detection algorithms RCNN, SPPNet, FastRCNN, FasterRCNN; target detection trick algorithms FPN, CascadeRCNN; deep learning one-stage target detection algorithms Yolo, X, SSD, RetinaNet; deep learning anchor-free target detection algorithms CornerNet, CenterNet, FCOS; Transformer-based target detection algorithm DETR, etc.

[0085] Preferably, this embodiment prefers the Yolo algorithm.

[0086] Step S3, classifying all plant organs through a classification algorithm, and obtaining an organ quantity based on each preset plant organ type.

[0087] Preferably, the preset plant organ type can be set according to the main organs of the plant, such as branches, leaves, flowers, and fruits. If soilless cultivation is adopted, the root system type can also be added.

[0088] Preferably, the classification algorithm of this embodiment is preferably Bayesian classification.

[0089] Specifically, Bayesian classification is a general term for a class of classification algorithms. These algorithms are all based on Bayesian theorem, so they are collectively called Bayesian classification. The Bayesian theorem is that let the sample space of experiment E be S, A be the event of E, is a partition of S, and , then Bayes' theorem is expressed as follows:

[0090] .

[0091] It should be noted that Bayes' theorem is a well-known theorem, and its symbolic meaning is not repeated here, and the symbolic meaning of the theorem is not interchangeable with the symbolic meanings of other formulas in this embodiment.

[0092] Preferably, the naive Bayes classification assumes that the existence of a specific feature in a class is independent of the existence of any other feature, that is, each feature is independent of each other. Therefore, there are some constraints on the actual situation. If there is a correlation between attributes, the classification accuracy will be reduced, but in actual application, the classification effect of naive Bayes is relatively accurate. Naive Bayes solves the probability of each category appearing under the condition that the item appears for a given item to be classified. The largest probability is considered to be the category to which the item to be classified belongs.

[0093] Specifically, Naive Bayes is defined as follows:

[0094] ① Let x={a1, a2, a3,…, an} be an item to be classified, and each a is a feature of x.

[0095] ②There is a category set c={y1, y2, y3,…, ym}.

[0096] ③Calculate P(y1|x), P(y2|x), …, P(ym|x).

[0097] ④If P(yk|x)=max{P(y1|x),P(y2|x),…,P(ym|x)}, then x∈yk.

[0098] Then calculate the conditional probabilities in step ③ through the following steps:

[0099] 1. Find a set of items to be classified with known classifications. This set is called a training sample set.

[0100] 2. Statistically obtain the conditional probability estimate of each feature attribute under each category. That is:

[0101] P(a1|y1),P(a2|y1),……,P(an|y1)

[0102] P(a1|y2),P(a2|y2),…,P(an|y2);

[0103] …

[0104] P(a1|ym),P(a2|ym),…,P(an|ym);

[0105] 3. Assuming that each feature attribute is conditionally independent, according to the Bayesian principle:

[0106] P(yi|x)=P(x|yi)P(yi) / p(x).

[0107] Since the denominator is a constant for all categories, we only need to maximize the numerator. And because each feature attribute is conditionally independent, then:

[0108] P(x|yi)P(yi)=P(a1|yi)P(a2|yi)…P(an|yi)P(yi).

[0109] It should be noted that the above preferred contents are also explanations of the principles, and the meanings of their symbols are not interchangeable with the meanings of the symbols of other formulas in this embodiment.

[0110] Step S4, respectively obtaining the image area value of each plant organ, and obtaining the sum of all the image area values.

[0111] Step S5, based on the same preset plant organ type, the organ quantity of all preset time periods is defined as an independent variable, the sum value is defined as a dependent variable, and the linear regression relationship between all independent variables and dependent variables is defined through a multiple linear regression model based on the same preset time period.

[0112] Preferably, in order to prevent the multivariate linear regression model from overfitting and causing the model's objective function to be inaccurate, the following method can be used to avoid this:

[0113] ① Increase the number of sample points in the training set.

[0114] ② Reduce the complexity of the model, that is, the model order.

[0115] ③ Regularization.

[0116] ④ Shorten the number of training rounds or lower the accuracy limit to stop training the model before the model overfits.

[0117] ⑤ Data enhancement and noise addition.

[0118] ⑥ Deep learning: Dropout and Dropconnect.

[0119] Step S6, calculating the regression coefficient of the linear regression relationship by the least square method and substituting it into the multivariate linear regression model to obtain an evaluation model.

[0120] Step S7, obtaining a real-time visual image of the green plant, and obtaining all real-time plant organs in the real-time visual image through a target detection algorithm.

[0121] Step S8, substituting all real-time plant organs into the evaluation model to obtain a real-time sum value.

[0122] Step S9, determining whether the real-time sum value is greater than the sum value, if so, determining that the green plant is in a healthy state.

[0123] Furthermore, in step S2, all plant organs in each visual image are obtained by using a target detection algorithm, including:

[0124] Step S21, dividing the current visual image into a plurality of grids on average.

[0125] Preferably, the number of grids can be set to 448×448, and the size of the image data can be adjusted to meet the specifications of 448×448, and then the resized image can be evenly divided into S×S (for example, 7×7) grids, and the size of each grid is 64×64.

[0126] Step S22, predicting a number of bounding boxes for all plant organs based on all grids according to a target detection algorithm.

[0127] Step S23, respectively obtain the confidence of each bounding box, obtain the bounding box with the largest confidence and mark it as the first-order bounding box.

[0128] Step S24, calculating the intersection-over-union ratio of each other bounding box with the first-order bounding box.

[0129] Step S25 , selecting a bounding box whose intersection-over-union ratio is greater than or equal to a preset threshold as a second-order bounding box.

[0130] Preferably, the intersection-over-union ratio is a ratio (which may be an area ratio) obtained by dividing the intersection of the second-order bounding box and each first-order bounding box by the union of the second-order bounding box and each first-order bounding box.

[0131] Step S26, obtaining a second-order bounding box with the highest confidence and defining it as a plant organ.

[0132] Preferably, each grid is used to predict The coordinates and width and height of the first-order bounding box, as well as the confidence of each first-order bounding box, that is, each grid needs to predict value.

[0133] It can be understood that each grid needs to predict indivual ;in, is the offset of the center of the first-order bounding box relative to the grid, is the ratio of the first-order bounding box to the above resized image, is the confidence of the grid, which takes a value of 1 or 0.

[0134] Preferably, the confidence level can be understood as whether there is a target in the current grid and the accuracy of the first-order bounding box.

[0135] For example: suppose there is an object in a resized image, and the width and height of the resized image are but:

[0136] Divide the image into 7×7 (S×S) grids evenly, and there is a grid located at the center of the target. The coordinates of the grid are , let the coordinates of the center of the target be , then according to Calculate the above offset.

[0137] Preferably, in actual detection, if the predicted first-order bounding box and the actual bounding box overlap perfectly, the intersection-and-union ratio is 1. In actual application, the value of the preset ratio can generally be set to 0.5 to determine whether the predicted second-order bounding box is correct, and the accuracy of the second-order bounding box is positively correlated with the intersection-and-union ratio.

[0138] Preferably, the YOLO algorithm also needs to train the first-order bounding box to improve the accuracy of target detection.

[0139] Next, the above training model is trained with a preset target training set, and the weights and biases of the training model are adjusted iteratively a certain number of times through the back propagation algorithm to reduce the value of the loss function of the training model.

[0140] Preferably, the loss function is .

[0141] in, For the Grid An indicator function of whether a first-order bounding box is responsible for the target, with a value of 1 or 0; , , , , Corresponding to indivual Predicted value.

[0142] It can be understood that the loss function includes the coordinate value deviation of the first-order bounding box, the confidence deviation, and the prediction probability deviation (or category deviation).

[0143] in, is the first-order bounding box midpoint loss in the coordinate value deviation, is the first-order bounding box width and height loss in the coordinate value deviation, is the confidence deviation, is the deviation of the predicted probability (or class deviation).

[0144] in, is the positioning error penalty. Generally, ; That is the S×S grids mentioned above; is the number of first-order bounding boxes; and For the The estimated values ​​of the horizontal and vertical coordinates of the midpoints of the first-order bounding boxes; and For the An estimate of the width and height of a first-order bounding box; For the The confidence of the first-order bounding box; For the An estimate of the confidence of a first-order bounding box; is the confidence prediction loss, generally ; For the The class probabilities of the first-order bounding boxes; For the An estimate of the class probability of a first-order bounding box; and In correspond .

[0145] It should be noted that since each grid does not necessarily contain a target, if there is no target in the grid, it will lead to The value of is 0, which makes the gradient span in the subsequent back propagation algorithm too large, so we introduce To control the loss of the first-order bounding box prediction position, and introduce Controls the penalty for non-existent targets within a single grid.

[0146] It should be noted that the symbolic meaning of the above YOLO algorithm is not interchangeable with the symbolic meanings of other positions in this embodiment.

[0147] Further, in step S3, all plant organs are classified by a classification algorithm, and an organ number is obtained based on each preset plant organ type, including:

[0148] Step S31: Pack all plant organs in the same preset time period into a data set to be classified , For the A data set to be classified for a preset time period, For the The first plant organs, is the number of all plant organs.

[0149] Step S32, defining a plant organ category set according to the number of preset plant organ types ,in, For category collection The preset plant organ types, is the number of preset plant organ types.

[0150] Step S33, calculate the first The conditional probability of each plant organ in the dataset to be classified under each preset plant organ type:

[0151] (1).

[0152] in, For the Preset plant organ type The first The conditional probability of a plant organ; For the The marginal probability of a preset plant organ type; For the Under the preset plant organ type, The first The conditional probability of a plant organ.

[0153] Step S34: All plant organs in the dataset to be classified for a preset time period are classified into the preset plant organ type with the highest conditional probability, and the first A classification table of plant organs for a preset time period.

[0154] Step S35, counting the number of organs under each preset plant organ type.

[0155] Further, step S5, based on the same preset plant organ type, the number of organs in all preset time periods is defined as an independent variable, the sum value is defined as a dependent variable, and the linear regression relationship between all independent variables and the dependent variable is defined based on the same preset time period through a multiple linear regression model, including:

[0156] Step S51, define the linear regression relationship between the dependent variable and all independent variables according to formula (2):

[0157] (2).

[0158] in, For the A dependent variable with a preset time period; is the intercept of the linear regression relationship; For the The first of the preset time periods Linear regression coefficients for each plant organ; is the random error of the linear regression relationship.

[0159] Further, in step S6, the regression coefficient of the linear regression relationship is calculated by the least square method and substituted into the multivariate linear regression model to obtain an evaluation model, including:

[0160] Step S61, calculate the first The regression coefficient of the linear regression relationship for a preset time period is:

[0161] (3).

[0162] in, for An estimated value of for The transposed matrix of .

[0163] Preferably, the square of the residual To test the significance of the regression model, if , it means that the model is more significant; if , it means that the significance of the model is general; if , it means that the significance of the model is poor, and the significance of the model is linearly positively correlated with the square of the residual.

[0164] Further, step S9, judging whether the real-time sum value is greater than the sum value, then comprising:

[0165] Step S10: If not, determine that the green plant is in an unhealthy state.

[0166] Step S20, obtaining the color data of the plant organs in each preset time period respectively through a target detection algorithm.

[0167] Step S30, obtaining the current average value of all color data of the same preset plant organ type in the current preset time period.

[0168] Step S40, obtaining the historical average value of all color data of the same preset plant organ type in the historical time period based on the preset backoff step number.

[0169] Step S50, determining the difference between the current average value and the historical average value.

[0170] Step S60: Mark the preset plant organ type whose difference is greater than or equal to the preset difference threshold as an unhealthy organ type.

[0171] Further, in step S60, the preset plant organ type with a difference greater than or equal to a preset difference threshold is marked as an unhealthy organ type, and then, the following steps are included:

[0172] Step S100, sending the unhealthy organ type to an external receiving end.

[0173] Preferably, since the withering characteristics of green plants are the transformation from green to yellow or black, the above-mentioned difference can be defined by identifying the RGB color value.

[0174] This embodiment obtains several visual images of green plants from at least two perspectives at intervals of preset time periods; obtains all plant organs in each visual image respectively through a target detection algorithm; classifies all plant organs through a classification algorithm, and obtains an organ number based on each preset plant organ type; obtains the image area value of each plant organ respectively, and obtains the sum of all image area values; defines the organ number of all preset time periods based on the same preset plant organ type as an independent variable, and defines the sum value as a dependent variable; defines the linear regression relationship between all independent variables and dependent variables based on the same preset time period through a multivariate linear regression model; calculates the regression coefficient of the linear regression relationship through the least squares method and substitutes it into the multivariate linear regression model to obtain an evaluation model; obtains real-time visual images of green plants, and obtains all real-time plant organs in the real-time visual images through a target detection algorithm; substitutes all real-time plant organs into the evaluation model to obtain a real-time sum value; determines whether the real-time sum value is greater than the sum value, and if so, determines that the green plants are in a healthy state. This embodiment utilizes the characteristic that the withering, falling off, wilting and death of plant leaves, branches, flowers and fruits are all shrinkage processes caused by water reduction. Its intuitive visual effect is the reduction of the visual area of ​​the plant under multiple viewing angles. The plant can show this characteristic when it is initially short of water. As the degree of water shortage increases, the visual area will continue to decrease, that is, the degree of water shortage and the visual area are positively correlated, so this embodiment can capture this characteristic through at least two viewing angles, and there is no need to identify the plant variety, and the health status can be judged only by the visual area.

[0175] like Figure 2 As shown, this embodiment provides an embodiment of a green plant health assessment device in a confined space. In this embodiment, the green plant health assessment device is applied to the green plant health assessment method as in the above-mentioned embodiment.

[0176] Specifically, the green plant health evaluation device includes a green plant visual image acquisition module 1, a plant organ acquisition module 2, a plant organ classification module 3, an image area value and sum value acquisition module 4, a linear regression relationship definition module 5, an evaluation model acquisition module 6, a real-time plant organ acquisition module 7, a real-time sum value acquisition module 8, and a real-time sum value judgment module 9, which are electrically connected in sequence.

[0177] Among them, the green plant visual image acquisition module 1 is used to acquire a number of visual images of green plants from at least two perspectives at intervals of a preset time period; the plant organ acquisition module 2 is used to respectively acquire all plant organs in each visual image through a target detection algorithm; the plant organ classification module 3 is used to classify all plant organs through a classification algorithm, and obtain an organ number based on each preset plant organ type; the image area value and sum value acquisition module 4 is used to respectively obtain the image area value of each plant organ, and obtain the sum of all image area values; the linear regression relationship definition module 5 is used to define the organ number of all preset time periods as an independent variable based on the same preset plant organ type, and the sum value is obtained. The sum is defined as the dependent variable, and the linear regression relationship between all independent variables and dependent variables is defined through a multiple linear regression model based on the same preset time period; the evaluation model acquisition module 6 is used to calculate the regression coefficient of the linear regression relationship through the least squares method and substitute it into the multiple linear regression model to obtain the evaluation model; the real-time plant organ acquisition module 7 is used to obtain the real-time visual image of the green plant, and obtain all the real-time plant organs of the real-time visual image through the target detection algorithm; the real-time sum value acquisition module 8 is used to substitute all the real-time plant organs into the evaluation model to obtain the real-time sum value; the real-time sum value judgment module 9 is used to judge whether the real-time sum value is greater than the sum value. If so, the green plant is judged to be in a healthy state.

[0178] Furthermore, the plant organ acquisition module 2 specifically includes a first plant organ acquisition submodule, a second plant organ acquisition submodule, a third plant organ acquisition submodule, a fourth plant organ acquisition submodule, a fifth plant organ acquisition submodule, and a sixth plant organ acquisition submodule, which are electrically connected in sequence; the first plant organ acquisition submodule is electrically connected to the green plant visual image acquisition module 1, and the sixth plant organ acquisition submodule is electrically connected to the plant organ classification module 3.

[0179] Among them, the first plant organ acquisition submodule is used to divide the current visual image into a number of grids on average; the second plant organ acquisition submodule is used to predict a number of bounding boxes for all plant organs based on all grids according to the target detection algorithm; the third plant organ acquisition submodule is used to obtain the confidence of each bounding box respectively, and obtain the bounding box with the largest confidence and mark it as a first-order bounding box; the fourth plant organ acquisition submodule is used to calculate the intersection and union ratio of each other bounding box with the first-order bounding box; the fifth plant organ acquisition submodule is used to select a bounding box with an intersection and union ratio greater than or equal to a preset threshold as a second-order bounding box; the sixth plant organ acquisition submodule is used to obtain the second-order bounding box with the highest confidence and define it as a plant organ.

[0180] Furthermore, the plant organ classification module 3 specifically includes a first plant organ classification submodule, a second plant organ classification submodule, a third plant organ classification submodule, a fourth plant organ classification submodule, and a fifth plant organ classification submodule, which are electrically connected in sequence; the first plant organ classification submodule is electrically connected to the sixth plant organ acquisition submodule, and the fifth plant organ classification submodule is electrically connected to the image area value and sum value acquisition module 4.

[0181] The first plant organ classification submodule is used to package all plant organs in the same preset time period into a data set to be classified. , For the A data set to be classified for a preset time period, For the The first plant organs, is the number of all plant organs.

[0182] The second plant organ classification submodule is used to define a plant organ category set according to the number of preset plant organ types. ,in, For category collection The preset plant organ types, is the number of preset plant organ types.

[0183] The third plant organ classification submodule is used to calculate the The conditional probability of each plant organ in the dataset to be classified under each preset plant organ type:

[0184] (1).

[0185] in, For the Preset plant organ type The first The conditional probability of a plant organ; For the The marginal probability of a preset plant organ type; For the Under the preset plant organ type, The first The conditional probability of a plant organ.

[0186] The fourth plant organ classification submodule is used to classify All plant organs in the dataset to be classified for a preset time period are classified into the preset plant organ type with the highest conditional probability, and the first A classification table of plant organs for a preset time period.

[0187] The fifth plant organ classification submodule is used to count the number of organs under each preset plant organ type.

[0188] Furthermore, the linear regression relationship definition module 5 is specifically used to define the linear regression relationship between the dependent variable and all independent variables according to formula (2):

[0189] (2).

[0190] in, For the A dependent variable with a preset time period; is the intercept of the linear regression relationship; For the The first of the preset time periods Linear regression coefficients for each plant organ; is the random error of the linear regression relationship.

[0191] Furthermore, the evaluation model acquisition module 6 is specifically used to calculate the first The regression coefficient of the linear regression relationship for a preset time period is:

[0192] (3).

[0193] in, for An estimated value of for The transposed matrix of .

[0194] Furthermore, the green plant health evaluation device also includes an unhealthy state determination module, a color data acquisition module, a color data current average value acquisition module, a color data historical average value acquisition module, a color data difference acquisition module, and an unhealthy organ type marking module, which are electrically connected in sequence; the unhealthy state determination module is electrically connected to the real-time sum value determination module 9.

[0195] Among them, the unhealthy state determination module is used to determine that the green plant is in an unhealthy state if the real-time sum is not greater than the sum; the color data acquisition module is used to obtain the color data of the plant organ in each preset time period through the target detection algorithm; the color data current average value acquisition module is used to obtain the current average value of all color data in the current preset time period under the same preset plant organ type; the color data historical average value acquisition module is used to obtain the historical average value of all color data in the historical time period based on a preset backoff step number under the same preset plant organ type; the color data difference acquisition module is used to determine the difference between the current average value and the historical average value; the unhealthy organ type marking module is used to mark the preset plant organ type whose difference is greater than or equal to the preset difference threshold as an unhealthy organ type.

[0196] Furthermore, the green plant health assessment device also includes an unhealthy organ type sending module electrically connected to the unhealthy organ type marking module.

[0197] The unhealthy organ type sending module is used to send the unhealthy organ type to an external receiving end.

[0198] It should be noted that this embodiment is a functional module embodiment based on the above method embodiment. The optimization, expansion, limitation, example, and principle description of this embodiment can be referred to the above embodiment, and this embodiment will not be repeated.

[0199] This embodiment obtains several visual images of green plants from at least two perspectives at intervals of preset time periods; obtains all plant organs in each visual image respectively through a target detection algorithm; classifies all plant organs through a classification algorithm, and obtains an organ number based on each preset plant organ type; obtains the image area value of each plant organ respectively, and obtains the sum of all image area values; defines the organ number of all preset time periods based on the same preset plant organ type as an independent variable, and defines the sum value as a dependent variable; defines the linear regression relationship between all independent variables and dependent variables based on the same preset time period through a multivariate linear regression model; calculates the regression coefficient of the linear regression relationship through the least squares method and substitutes it into the multivariate linear regression model to obtain an evaluation model; obtains real-time visual images of green plants, and obtains all real-time plant organs in the real-time visual images through a target detection algorithm; substitutes all real-time plant organs into the evaluation model to obtain a real-time sum value; determines whether the real-time sum value is greater than the sum value, and if so, determines that the green plants are in a healthy state. This embodiment utilizes the characteristic that the withering, falling off, wilting and death of plant leaves, branches, flowers and fruits are all shrinkage processes caused by water reduction. Its intuitive visual effect is the reduction of the visual area of ​​the plant under multiple viewing angles. The plant can show this characteristic when it is initially short of water. As the degree of water shortage increases, the visual area will continue to decrease, that is, the degree of water shortage and the visual area are positively correlated, so this embodiment can capture this characteristic through at least two viewing angles, and there is no need to identify the plant variety, and the health status can be judged only by the visual area.

[0200] Figure 3 Schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Figure 3 As shown, the electronic device 10 includes a processor 101 and a memory 102 coupled to the processor 101 .

[0201] The memory 102 stores program instructions for implementing a method for evaluating the health of green plants in a confined space according to any one of the above embodiments.

[0202] The processor 101 is used to execute program instructions stored in the memory 102 to perform health assessment of green plants in a confined space.

[0203] The processor 101 may also be referred to as a CPU (Central Processing Unit). The processor 101 may be an integrated circuit chip having a signal processing capability. The processor 101 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0204] Further, Figure 4 This is a schematic diagram of the structure of a storage medium according to an embodiment of the present application. Figure 4 The storage medium 11 of the embodiment of the present application stores program instructions 111 that can implement all the above methods, wherein the program instructions 111 can be stored in the above storage medium in the form of a software product, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes, or terminal devices such as computers, servers, mobile phones, tablets, etc.

[0205] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0206] In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. The above is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of this application, or directly or indirectly used in other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A method for evaluating the health of green plants in a confined space, wherein the confined space is enclosed by four transparent baffles covered by a cover plate and a matrix in sequence from top to bottom, and green plants are planted on the matrix, characterized in that: The green plant health evaluation method comprises: Step S1, acquiring a plurality of visual images of the green plant from at least two viewing angles at intervals of a preset time period; Step S2, obtaining all plant organs in each visual image respectively through a target detection algorithm; Step S3, classifying all plant organs by a classification algorithm, and obtaining an organ number based on each preset plant organ type; Step S4, respectively obtaining the image area value of each plant organ, and obtaining the sum of all the image area values; Step S5, defining the number of organs in all preset time periods based on the same preset plant organ type as an independent variable, defining the sum value as a dependent variable, and defining the linear regression relationship between all independent variables and the dependent variable based on the same preset time period through a multiple linear regression model; Step S6, calculating the regression coefficient of the linear regression relationship by the least square method and substituting it into the multivariate linear regression model to obtain an evaluation model; Step S7, obtaining a real-time visual image of the green plant, and obtaining all real-time plant organs in the real-time visual image through the target detection algorithm; Step S8, substituting all real-time plant organs into the evaluation model to obtain a real-time sum value; Step S9, determining whether the real-time sum value is greater than the sum value, and if so, determining that the green plant is in a healthy state.

2. The green plant health evaluation method according to claim 1, characterized in that: Step S2, obtaining all plant organs in each visual image by using a target detection algorithm, including: Step S21, dividing the current visual image into a plurality of grids on average; Step S22, predicting a number of bounding boxes for all plant organs based on all grids according to the target detection algorithm; Step S23, respectively obtaining the confidence of each bounding box, obtaining the bounding box with the largest confidence and marking it as a first-order bounding box; Step S24, calculating the intersection-over-union ratio of each other bounding box with the first-order bounding box; Step S25, selecting a bounding box whose intersection-over-union ratio is greater than or equal to a preset threshold as a second-order bounding box; Step S26, obtaining a second-order bounding box with the highest confidence and defining it as the plant organ.

3. The green plant health evaluation method according to claim 1, characterized in that: Step S3, classifying all plant organs by a classification algorithm, and obtaining an organ quantity based on each preset plant organ type, including: Step S31: Pack all plant organs in the same preset time period into a data set to be classified , For the A data set to be classified for a preset time period, For the The first plant organs, is the number of all plant organs; Step S32, defining a plant organ category set according to the number of the preset plant organ types ,in, For the category set The preset plant organ types, is the number of the preset plant organ types; Step S33, calculate the first The conditional probability of each plant organ in the dataset to be classified under each preset plant organ type: （1）； in, For the Preset plant organ type The first The conditional probability of a plant organ; For the The marginal probability of a preset plant organ type; For the Under the preset plant organ type, The first The conditional probability of a plant organ; Step S34: All plant organs in the data set to be classified for a preset time period are classified into the preset plant organ type with the highest conditional probability, and the first A classification table of plant organs for a preset time period; Step S35, counting the number of organs under each preset plant organ type.

4. The green plant health evaluation method according to claim 3, characterized in that: Step S5, based on the same preset plant organ type, defining the number of organs in all preset time periods as an independent variable, defining the sum value as a dependent variable, and defining the linear regression relationship between all independent variables and the dependent variable based on the same preset time period through a multiple linear regression model, including: Step S51, defining the linear regression relationship between the dependent variable and all independent variables according to formula (2): （2）； in, For the A dependent variable with a preset time period; is the intercept of the linear regression relationship; For the The first of the preset time periods Linear regression coefficients for each plant organ; is the random error of the linear regression relationship.

5. The green plant health evaluation method according to claim 4, characterized in that: Step S6, calculating the regression coefficient of the linear regression relationship by the least square method and substituting it into the multivariate linear regression model to obtain an evaluation model, including: Step S61, calculate the first The regression coefficient of the linear regression relationship for a preset time period is: （3）； in, for An estimated value of for The transposed matrix of .

6. The green plant health evaluation method according to claim 1, characterized in that: Step S9, judging whether the real-time sum value is greater than the sum value, and then comprising: Step S10, if not, determining that the green plant is in an unhealthy state; Step S20, respectively acquiring color data of the plant organ in each preset time period by using the target detection algorithm; Step S30, obtaining the current average value of all color data of the same preset plant organ type in the current preset time period; Step S40, obtaining the historical average value of all color data of the same preset plant organ type in the historical time period based on the preset backoff step number; Step S50, determining the difference between the current average value and the historical average value; Step S60: marking the preset plant organ type whose difference is greater than or equal to the preset difference threshold as an unhealthy organ type.

7. The green plant health evaluation method according to claim 6, characterized in that: Step S60, marking the preset plant organ type whose difference is greater than or equal to the preset difference threshold as an unhealthy organ type, and then comprising: Step S100, sending the unhealthy organ type to an external receiving end.

8. A green plant health assessment device in a confined space, the green plant health assessment device being applied to the green plant health assessment method according to any one of claims 1 to 7, characterized in that: The green plant health evaluation device comprises: A green plant visual image acquisition module, used to acquire a plurality of visual images of the green plant from at least two viewing angles at intervals of a preset time period; A plant organ acquisition module is used to acquire all plant organs in each visual image through a target detection algorithm; A plant organ classification module, used to classify all plant organs through a classification algorithm, and obtain an organ number based on each preset plant organ type; An image area value and sum value acquisition module is used to respectively acquire the image area value of each plant organ and acquire the sum value of all image area values; A linear regression relationship definition module, used to define the number of organs in all preset time periods based on the same preset plant organ type as an independent variable, define the sum value as a dependent variable, and define the linear regression relationship between all independent variables and the dependent variable based on the same preset time period through a multivariate linear regression model; An evaluation model acquisition module, used to calculate the regression coefficient of the linear regression relationship by the least square method and substitute it into the multivariate linear regression model to obtain an evaluation model; A real-time plant organ acquisition module, used to acquire the real-time visual image of the green plant, and acquire all real-time plant organs of the real-time visual image through the target detection algorithm; A real-time sum value acquisition module is used to substitute all real-time plant organs into the evaluation model to obtain a real-time sum value; The real-time sum value judgment module is used to judge whether the real-time sum value is greater than the sum value, and if so, it is judged that the green plant is in a healthy state.

9. An electronic device, characterized in that: It includes a processor and a memory coupled to the processor, wherein the memory stores program instructions that can be executed by the processor; when the processor executes the program instructions stored in the memory, the green plant health assessment method as described in any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium stores program instructions, and when the program instructions are executed by the processor, the green plant health assessment method as described in any one of claims 1 to 7 can be implemented.

Images