A method and system for identifying the number and position of missing seedlings of a seedling

Abstract

The present invention discloses a method and a system for identifying the number and position of missing seedlings, which relate to the field of identifying the number and position of missing seedlings in paddy fields. The method includes: using a drone carrying an RGB camera to collect image information of a target area for missing seedling detection; preprocessing the image information to obtain an HSV color space model; segmenting the HSV color space model by using an empirical threshold method to obtain a binary image of the seedling image; determining the root position of the seedlings according to the binary image of the seedling image; performing linear regression fitting by using the root position of the seedlings to obtain a fitting straight line of the seedlings; calculating the number of missing seedlings according to the fitting straight line of the seedlings; and determining the coordinates of the position of the missing seedling on the fitting straight line of the seedlings. The present invention can quickly and accurately detect the number and position of missing seedlings, reduce the labor intensity and operation cost, and improve the operation efficiency and operation accuracy.

Description

Technical Field

[0001] The present invention relates to the field of identifying the number and position of missing rice seedlings in paddy fields, and particularly to a method and system for identifying the number and position of missing rice seedlings. Background Art

[0002] The quality of rice transplanting and the number of missing seedlings have a great impact on the final yield of rice. Traditional methods for detecting the number and position of missing seedlings in rice transplanting are labor-intensive, costly, and subjective. They mainly rely on manual field inspection of the cutting situation of each row of rice seedlings to determine the number and position of missing seedlings. With the increase in the number of large farms and the sharp rise in labor costs, rice growers have further raised their requirements for the accuracy and speed of detecting the number and position of missing seedlings in rice transplanting. How to improve the speed and accuracy of detecting the number and position of missing seedlings in rice transplanting is an urgent problem to be solved currently. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and system for identifying the number and position of missing rice seedlings, which can quickly and accurately detect the number and position of missing seedlings, reduce labor intensity and operation costs, and improve operation efficiency and operation accuracy.

[0004] To achieve the above purpose, the present invention provides the following solutions:

[0005] A method for identifying the number and position of missing rice seedlings, the method comprising:

[0006] Collecting image information of the target area for missing seedling detection by using a drone carrying an RGB camera;

[0007] Preprocessing the image information to obtain an HSV color space model;

[0008] Segmenting the HSV color space model by using an empirical threshold method to obtain a binary image of the rice seedling image;

[0009] Determining the root position of the rice seedlings according to the binary image of the rice seedling image;

[0010] Performing linear regression fitting by using the root position of the rice seedlings to obtain a fitting straight line of the rice seedlings;

[0011] Calculating the number of missing rice seedlings according to the fitting straight line of the rice seedlings;

[0012] Determining the coordinates of the missing seedling position on the fitting straight line of the rice seedlings according to the number of missing rice seedlings.

[0013] Optionally, the preprocessing the image information to obtain an HSV color space model specifically includes:

[0014] Perform image stitching on the said image information;

[0015] Perform orthophoto rectification and multispectral image spectral correction on the stitched image information to obtain the RGB orthophoto of the seedling missing detection target area;

[0016] Convert the said RGB orthophoto into the HSV color space model.

[0017] Optionally, the determination of the root position of the seedling according to the binary map of the seedling image specifically includes:

[0018] Extract the binary image of the seedling skeleton according to the binary map of the seedling image;

[0019] Remove redundant branches in the binary image of the seedling skeleton to obtain the seedling morphology map;

[0020] Use the Harris corner detection algorithm to perform corner detection on the seedling morphology map to determine the root position of the seedling.

[0021] Optionally, the calculation of the number of missing seedlings according to the fitted straight line of the seedling specifically includes:

[0022] Calculate the projection point coordinates of the root position of the seedling on the corresponding fitted straight line of the seedling;

[0023] According to the projection point coordinates, calculate the distance between the projections of adjacent two seedlings in each row within the seedling missing detection target area, the distance between the actual starting position and the set starting position of the seedlings in each row within the seedling missing detection target area, and the average distance of the seedling projections;

[0024] When the distance between the projections of adjacent two seedlings is greater than the set distance, the area between the adjacent two seedlings is the middle seedling missing area; the set distance is the set multiple of the average distance of the seedling projections;

[0025] When the distance between the actual starting position and the set starting position of the seedling is greater than the set distance, the area between the actual starting position and the set starting position of the seedling is the starting seedling missing area;

[0026] According to the average distance of the seedling projections, calculate the number of missing seedlings in the middle seedling missing area and the number of missing seedlings in the starting seedling missing area to obtain the number of missing seedlings of the seedling.

[0027] Optionally, the determination of the coordinates of the missing seedling position on the fitted straight line of the seedling specifically includes:

[0028] According to the number of missing seedlings of the seedling and the average distance of the seedling projections, determine the projection point coordinates of the root position of the missing seedling on the corresponding fitted straight line of the seedling;

[0029] The coordinates of the seedlings at the missing seedling position on the fitted straight line of the seedlings are determined according to the coordinates of the projection points.

[0030] A system for identifying the number and location of missing seedlings, applied to the above-mentioned method for identifying the number and location of missing seedlings, comprises:

[0031] The acquisition module is used to use a drone equipped with an RGB camera to collect image information of the target area for leaking rice seedlings detection;

[0032] A preprocessing module, used for preprocessing the image information to obtain an HSV color space model;

[0033] A segmentation module is used to segment the HSV color space model using an empirical threshold method to obtain a binary image of the seedling image;

[0034] a root position determination module, configured to determine the root position of the seedling according to the binary image of the seedling image;

[0035] A fitting module, configured to perform linear regression fitting using the root position of the seedling to obtain a fitting straight line of the seedling;

[0036] A calculation module, configured to calculate the number of missing seedlings based on the fitted straight line of the seedlings;

[0037] The coordinate determination module is used to determine the coordinates of the missing seedling positions on the fitted straight line of the seedlings according to the number of missing seedlings.

[0038] Optionally, the preprocessing module includes:

[0039] A stitching submodule, configured to stitch the image information;

[0040] The correction submodule is used to perform orthophoto correction and multispectral image spectral correction on the spliced image information to obtain the RGB orthophoto image of the target area for leaking rice seedlings detection;

[0041] The conversion submodule is used to convert the RGB orthophoto image into an HSV color space model.

[0042] Optionally, the root position determination module includes:

[0043] an extraction submodule, configured to extract a binary image of the seedling skeleton based on the binary image of the seedling image;

[0044] a removal submodule, configured to remove redundant branches from the binary image of the rice seedling skeleton to obtain a rice seedling morphology image;

[0045] A detection sub-module, which is used to detect the corners of the seedling morphology map by using the Harris corner detection algorithm to determine the root position of the seedlings.

[0046] Optionally, the calculation module includes:

[0047] A projection point coordinate calculation sub-module, which is used to calculate the projection point coordinates of the root position of the seedling on the fitted straight line of the corresponding seedling;

[0048] A spacing calculation sub-module, which is used to calculate the spacing between the projections of adjacent two seedlings in each row within the target area for missing seedling detection, the spacing between the actual starting position and the set starting position of the seedlings in each row within the target area for missing seedling detection, and the average spacing of the seedling projections according to the projection point coordinates;

[0049] A first comparison sub-module, which is used to determine that there is a missing seedling in the middle between adjacent two seedlings when the spacing between the projections of adjacent two seedlings is greater than the set distance; the set distance is the set multiple of the average spacing of the seedling projections;

[0050] A second comparison sub-module, which is used to determine that there is a missing seedling at the start between the actual starting position and the set starting position of the seedling when the spacing between the actual starting position and the set starting position of the seedling is greater than the set distance;

[0051] A missing seedling quantity determination sub-module, which is used to calculate the quantity of missing seedlings at the middle missing seedling position and the quantity of missing seedlings at the start missing seedling position according to the average spacing of the seedling projections to obtain the quantity of missing seedlings of the seedlings.

[0052] Optionally, the coordinate determination module includes:

[0053] A missing seedling projection point coordinate determination sub-module, which is used to determine the projection point coordinates of the root position of the missing seedling on the fitted straight line of the corresponding seedling according to the quantity of missing seedlings of the seedling and the average spacing of the seedling projections;

[0054] A missing seedling coordinate determination sub-module, which is used to determine the coordinates of the seedling at the missing seedling position on the fitted straight line of the seedling according to the projection point coordinates.

[0055] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0056] A method for identifying the number and position of missing seedlings of seedlings provided by the present invention includes: using a drone carrying an RGB camera to collect image information of the target area for detecting missing seedlings; preprocessing the image information to obtain an HSV color space model; segmenting the HSV color space model by using an empirical threshold method to obtain a binary image of the seedling image; determining the root position of the seedling according to the binary image of the seedling image; performing linear regression fitting by using the root position of the seedling to obtain a fitting straight line of the seedling; calculating the number of missing seedlings of the seedling according to the fitting straight line of the seedling; and determining the coordinates of the position of the missing seedling on the fitting straight line of the seedling according to the number of missing seedlings of the seedling. The present invention uses a drone to collect the target area for detecting missing seedlings of rice transplanting, and uses an image morphology processing method and an image segmentation processing method to quickly and accurately detect the number and position of missing seedlings, reduce the labor intensity and operation cost, and improve the operation efficiency and operation accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the following described drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0058] Figure 1 It is a flow chart of the method for identifying the number and position of missing seedlings of seedlings provided by the present invention;

[0059] Figure 2 It is an orthoimage effect diagram of RGB provided by the present invention;

[0060] Figure 3 It is an effect diagram of the seedling identification result provided by the present invention;

[0061] Figure 4 It is an effect diagram of the missing seedling positioning of the first row provided by the present invention;

[0062] Figure 5 It is an enlarged effect diagram of the processing result of the first row provided by the present invention;

[0063] Figure 6 It is an effect diagram of the processing results of seedling detection and missing seedling detection provided by the present invention;

[0064] Figure 7 It is a module diagram of the system for identifying the number and position of missing seedlings of seedlings provided by the present invention.

[0065] Corresponding description:

[0066] 1 - Acquisition module, 2 - Pre - processing module, 3 - Segmentation module, 4 - Root position determination module, 5 - Fitting module, 6 - Calculation module, 7 - Coordinate determination module. Specific implementation manner

[0067] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0068] The purpose of the present invention is to provide a method and system for identifying the number and position of missing seedlings of seedlings, which can quickly and accurately detect the number and position of missing seedlings, reduce the labor intensity and operation cost, and improve the operation efficiency and operation accuracy.

[0069] To make the above - mentioned objects, features and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific implementation manners.

[0070] As Figure 1 shown, a method for identifying the number and position of missing seedlings of seedlings provided by the present invention includes:

[0071] Step S1: Use a drone equipped with an RGB camera to collect image information of the target area for missing - seedling detection; in practical applications, use a drone equipped with an RGB camera to collect image information of the target area for missing - seedling detection. The aerial multi - light camera is arranged in the direction of the drone's heading and set to follow the heading. When collecting image information, the self - stabilizing gimbal ensures that the lens direction is perpendicular to the ground; when the aerial camera is arranged in the direction of the drone's heading and set to follow the heading, the drone's heading repetition rate is set to not less than 60%, the drone's lateral repetition rate is set to not less than 55%, the flight altitude is set to 50 m, and the flight speed is not greater than 5 m / s.

[0072] Step S2: Pre - process the image information to obtain an HSV color - space model.

[0073] S2 specifically includes:

[0074] Step S21: Perform image stitching on the image information; specifically, complete image stitching according to the same features of the overlapping parts of the photos in the front - back direction of the heading and the left - right direction of the lateral.

[0075] Step S22: Perform orthophoto rectification and multispectral image spectral correction on the stitched image information to obtain an RGB orthophoto image of the target area for missing - seedling detection; specifically, with Figure 2Taking the first line in [ID] as an example, the orthophoto correction and multispectral image spectral correction are performed on the spliced image information to obtain the RGB orthophoto of the target area for missing seedling detection.

[0076] Step S23: Convert the RGB orthophoto into the HSV color space model. Specifically, as Figure 3 shown, in order to obtain accurate shape information of the seedlings, the original RGB orthophoto is converted into the HSV color space model to enhance the UAV image, that is, to perform image enhancement on the RGB orthophoto.

[0077] Step S3: Segment the HSV color space model using the empirical threshold method to obtain the binary image of the seedling image.

[0078] Specifically, find the maximum value MaxValue and minimum value MinValue of the seedling pixels in the HSV color space model image (image size is M*N), and use the empirical threshold method to adjust the optimal threshold range. The image is divided into foreground (seedlings) and background (soil background), and the binary image of the seedling image of the seedlings in the field is segmented. The decision formula of the threshold segmentation model is as follows:

[0079] B i = 0 (Pi < MinValue or P i > MaxValue);

[0080] F i = 1 (MinValue ≤ P i ≤ MaxValue);

[0081] where, B i is the background pixel value, F i is the foreground pixel value, P i is the value of the pixel of the HSV color model image, and i is the pixel serial number, where 1 ≤ i ≤ M*N.

[0082] Step S4: Determine the root position of the seedlings according to the binary image of the seedling image.

[0083] S4 specifically includes:

[0084] Step S41: Extract the binary image of the seedling skeleton according to the binary image of the seedling image.

[0085] Specifically, according to the binary image of the seedling image, the seedling skeleton is extracted. The skeleton structure is an important topological description of the image, which can visually retain the original shape of the object and is convenient for describing the morphology of the target and extracting features.

[0086] Step S42: Remove redundant branches in the binary image of the seedling skeleton to obtain a seedling morphology map. Specifically, by removing redundant branches from the skeleton binary image, a rice seedling morphology map with higher accuracy is obtained.

[0087] Step S43: Use the Harris corner detection algorithm to detect corners in the seedling morphology map and determine the root position of the seedling.

[0088] Specifically, the intersection point of the seedling leaves is the root position of the seedling plant. Use the Harris corner detection algorithm to detect corners on the seedling skeleton, and the detected corners are the leaf intersection points (X center , Y center ), Figure 4 is the positioning effect diagram of missing seedlings in the first row, as shown in Figure 4 shown.

[0089] Step S5: Use the root position of the seedling for linear regression fitting to obtain the fitted straight line of the seedling.

[0090] In practical applications, as shown in Figure 5 , the seedlings in each row are basically planted in a straight line along the ridge. It can be considered that there is a linear relationship between the central point coordinates X center and Y center . Use the central point coordinates of the seedlings in each row for linear regression fitting to obtain the planting straight line of each row in each channel image of the fusion image. For a row of seedlings X = (x1, x2,..., x n ), the linear relationship with y can be expressed as: y = β0 + β1x1 +... + β n x n + ε.

[0091] For the fitted straight lines of different rows of seedlings, Y = (y1, y2,..., y n ) can be obtained, then: Y = Xβ + ε.

[0092] Step S6: Calculate the number of missing seedlings of the seedling according to the fitted straight line of the seedling.

[0093] S6 specifically includes:

[0094] Step S61: Calculate the projection point coordinates of the root position of the seedling on the corresponding fitted straight line of the seedling. Specifically, the distance between the seedlings in each row is calculated through the projection points of the plant centers to the regression straight lines of the corresponding rows. The calculation formula is:

[0095] The projection point coordinates of the seedling plant are obtained as (X projection , Y projection ), where β = (β0, β1,..., β n ) is the coefficient.

[0096] Step S62: Calculate the spacing between the projections of two adjacent seedlings in each row within the target area for detecting missing seedlings, the spacing between the actual starting position and the set starting position of the seedlings in each row within the target area for detecting missing seedlings, and the average spacing of the projections of the seedlings.

[0097] In practical applications, select the image of the area where there is no missing seedling, repeat Step S3, Step S4, Step S5, and Step S61 to obtain the projection point coordinates (X projection , Y projection ) of each row of seedling plants, and calculate the average spacing of the projections of the seedling plants. The formula is:

[0098]

[0099] where (X projection1 , Y projection1 ) and (X projection2 , Y projection2 ) are the coordinates of two adjacent seedling plants, distance is the spacing between the central projection points of the two seedlings, is the average plant spacing of the projection, and number(plant) is the number of plants.

[0100] Step S63: When the spacing between the projections of two adjacent seedlings is greater than the set distance, the area between the two adjacent seedlings is the middle missing-seedling area; the set distance is the set multiple of the average spacing of the projections of the seedlings.

[0101] Specifically, set a set multiple k during the calculation process, that is, when the spacing between the central projections of adjacent plants is greater than k times the average plant spacing of the projection it is considered as the missing-seedling area. The formula for calculating the number of missing seedlings is:

[0102] where m is the number of missing seedlings at the place where the missing-seedling phenomenon occurs.

[0103] Step S64: When the spacing between the actual starting position and the set starting position of the seedlings is greater than the set distance, the area between the actual starting position and the set starting position of the seedlings is the starting missing-seedling area.

[0104] Specifically, for each row, the theoretical starting seedling Y coordinate is defined as the minimum Y min coordinate of the seedlings in the whole field. The actual starting seedling of each row is recorded as Y ture , when it is considered that the missing-seedling phenomenon occurs. The formula for calculating the missing seedlings at the start of each row is:

[0105] where n is the number of missing seedlings at the start of each row.

[0106] Step S65: Calculate the number of missing seedlings at the middle missing-seedling position and the number of missing seedlings at the starting missing-seedling position according to the average spacing of the seedling projections, so as to obtain the number of missing seedlings of the seedlings.

[0107] Specifically, calculate the total number of missing seedlings in all rows of the field. The formula is:

[0108] total number = sum(m) + sum(n); where total number is the total number of missing seedlings.

[0109] Step S7: Determine the coordinates of the missing-seedling position on the fitting straight line of the seedlings according to the number of missing seedlings of the seedlings.

[0110] S7 specifically includes:

[0111] Step S71: Determine the coordinates of the projection point of the root position of the missing-seedling seedlings on the corresponding fitting straight line of the seedlings according to the number of missing seedlings of the seedlings and the average spacing of the seedling projections.

[0112] Step S72: Determine the coordinates of the seedlings at the missing-seedling position on the fitting straight line of the seedlings according to the coordinates of the projection point.

[0113] Specifically, according to the number of missing seedlings, the range where the missing-seedling phenomenon occurs, and the average plant spacing, the coordinates of the corresponding missing-seedling position on the fitting straight line can be calculated. The specific calculation formula is:

[0114] where i is the serial number of the plant with the missing-seedling phenomenon in this row, and j is the serial number of the missing plant.

[0115] Applying the method for identifying the number of missing seedlings and the missing-seedling position of the seedlings provided by the present invention to the entire field, the detection results are as Figure 6 shown; where the X and Y coordinates are in pixels, the hollow circles are the centers of the plants, the dashed lines are the ridge division lines, and the straight line is the first-row linear regression fitting straight line.

[0116] An identification system for the number of missing seedlings and the missing-seedling position of the seedlings provided by the present invention is applied to the method for identifying the number of missing seedlings and the missing-seedling position of the seedlings as described above, as Figure 7 shown. The system includes:

[0117] An acquisition module 1, configured to acquire image information of a missing-seedling detection target area by using a drone carrying an RGB camera.

[0118] A preprocessing module 2, configured to preprocess the image information to obtain an HSV color space model.

[0119] A segmentation module 3, which is used to segment the HSV color space model by using an empirical threshold method to obtain a binary image of the seedling image.

[0120] A root position determination module 4, which is used to determine the root position of the seedling according to the binary image of the seedling image.

[0121] A fitting module 5, which is used to perform linear regression fitting by using the root position of the seedling to obtain a fitting straight line of the seedling.

[0122] A calculation module 6, which is used to calculate the number of missing seedlings of the seedling according to the fitting straight line of the seedling.

[0123] A coordinate determination module 7, which is used to determine the coordinates of the missing seedling position on the fitting straight line of the seedling according to the number of missing seedlings of the seedling.

[0124] As a specific implementation manner, the preprocessing module 2 includes:

[0125] A splicing sub-module, which is used to perform image splicing on the image information.

[0126] A rectification sub-module, which is used to perform orthophoto rectification and multispectral image spectral rectification on the spliced image information to obtain an RGB orthophoto image of the missing seedling detection target area.

[0127] A conversion sub-module, which is used to convert the RGB orthophoto image into an HSV color space model.

[0128] As a specific implementation manner, the root position determination module 4 includes:

[0129] An extraction sub-module, which is used to extract a binary image of the seedling skeleton according to the binary image of the seedling image.

[0130] A removal sub-module, which is used to remove redundant branches in the binary image of the seedling skeleton to obtain a seedling morphology image.

[0131] A detection sub-module, which is used to perform corner detection on the seedling morphology image by using the Harris corner detection algorithm to determine the root position of the seedling.

[0132] As a specific implementation manner, the calculation module 6 includes:

[0133] A projection point coordinate calculation sub-module, which is used to calculate the projection point coordinates of the root position of the seedling on the corresponding fitting straight line of the seedling.

[0134] A spacing calculation sub-module, configured to calculate the spacing between adjacent two seedling projections in each row within the target area for missing seedling detection, the spacing between the actual starting position and the set starting position of the seedlings in each row within the target area for missing seedling detection, and the average spacing of the seedling projections according to the coordinates of the projection points.

[0135] A first comparison sub-module, configured to determine that there is a missing seedling in the middle between adjacent two seedlings when the spacing between the adjacent two seedling projections is greater than a set distance; the set distance is a set multiple of the average spacing of the seedling projections.

[0136] A second comparison sub-module, configured to determine that there is a missing seedling at the starting position between the actual starting position and the set starting position of the seedling when the spacing between the actual starting position and the set starting position of the seedling is greater than a set distance.

[0137] A missing seedling quantity determination sub-module, configured to calculate the quantity of missing seedlings at the middle missing seedling position and the quantity of missing seedlings at the starting missing seedling position according to the average spacing of the seedling projections, so as to obtain the quantity of missing seedlings of the seedlings.

[0138] In addition, the coordinate determination module 7 includes:

[0139] A missing seedling projection point coordinate determination sub-module, configured to determine the projection point coordinates of the root position of the missing seedling on the fitting straight line of the corresponding seedling according to the quantity of missing seedlings of the seedlings and the average spacing of the seedling projections.

[0140] A missing seedling coordinate determination sub-module, configured to determine the coordinates of the seedling at the missing seedling position on the fitting straight line of the seedlings according to the projection point coordinates.

[0141] A method and system for identifying the quantity and position of missing seedlings of seedlings provided by the present invention have some advantages:

[0142] (1) Low labor intensity and operation cost, and high operation efficiency.

[0143] For traditional rice transplanting missing seedling detection, it is necessary to manually check the cutting situation of each row of seedlings in the paddy field on site to determine the quantity and position of missing seedlings, which has a large labor intensity, high cost and long time consumption. However, the present invention uses an unmanned aerial vehicle to collect the target area for rice transplanting missing seedling detection, and uses the image morphology processing method and the image segmentation processing method to quickly and accurately detect the quantity and position of missing seedlings, with high speed and short time. It is not necessary for personnel to enter the paddy field to detect the quantity and position of missing seedlings, and the operation is convenient and efficient.

[0144] (2) High precision and stability.

[0145] The traditional method for detecting missing seedlings in rice transplanting is highly subjective and mainly relies on the professional knowledge of operators. The differences in the professional knowledge and experience of operators lead to inconsistent measurement accuracies, making it difficult to maintain stability and credibility in the detection of missing seedlings in rice transplanting. The method of the present invention is mature and stable, with a fixed and streamlined detection process, excluding human influence. Through actual production verification, the accuracy can reach over 98.5%, and the stability can reach over 99%.

[0146] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments, and the same or similar parts among the various embodiments can be referred to each other. For the system disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the description in the method section.

[0147] Specific examples are used in this article to elaborate on the principles and implementation methods of the present invention. The descriptions of the above embodiments are only used to help understand the method of the present invention and its core idea; at the same time, for those of ordinary skill in the art, there will be changes in the specific implementation methods and application scopes according to the idea of the present invention. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for identifying the number and position of missing seedlings of rice seedlings, characterized in that, The method for identifying the missing seedling quantity and position of the seedlings includes: Using a drone equipped with an RGB camera to collect image information of the target area for missing seedling detection; Preprocessing the image information to obtain an HSV color space model; Segmenting the HSV color space model using an empirical threshold method to obtain a binary image of the seedling image; Determining the root position of the seedlings according to the binary image of the seedling image; Performing linear regression fitting using the root position of the seedlings to obtain a fitting straight line of the seedlings; Calculating the missing seedling quantity of the seedlings according to the fitting straight line of the seedlings; Determining the coordinates of the missing seedling position on the fitting straight line of the seedlings according to the missing seedling quantity of the seedlings; The step of calculating the missing seedling quantity of the seedlings according to the fitting straight line of the seedlings specifically includes: Calculating the projection point coordinates of the root position of the seedlings on the corresponding fitting straight line of the seedlings; According to the projection point coordinates, calculating the spacing between the projections of adjacent two seedlings in each row within the target area for missing seedling detection, the spacing between the actual starting position and the set starting position of the seedlings in each row within the target area for missing seedling detection, and the average spacing of the seedling projections; When the spacing between the projections of adjacent two seedlings is greater than the set distance, the area between the adjacent two seedlings is the middle missing seedling area; the set distance is a set multiple of the average spacing of the seedling projections; When the spacing between the actual starting position and the set starting position of the seedlings is greater than the set distance, the area between the actual starting position and the set starting position of the seedlings is the starting missing seedling area; Calculating the missing seedling quantity in the middle missing seedling area and the missing seedling quantity in the starting missing seedling area according to the average spacing of the seedling projections to obtain the missing seedling quantity of the seedlings; The step of determining the coordinates of the missing seedling position on the fitting straight line of the seedlings according to the missing seedling quantity of the seedlings specifically includes: Determining the projection point coordinates of the root position of the missing seedling on the corresponding fitting straight line of the seedlings according to the missing seedling quantity of the seedlings and the average spacing of the seedling projections; Determining the coordinates of the missing seedling on the fitting straight line of the seedlings according to the projection point coordinates; 2. The method for identifying the missing seedling quantity and missing seedling position of the seedlings according to claim 1, characterized in that The step of preprocessing the image information to obtain an HSV color space model specifically includes: Performing image stitching on the image information; Performing orthophoto rectification and multispectral image spectral correction on the stitched image information to obtain an RGB orthophoto image of the target area for missing seedling detection; Converting the RGB orthophoto image into an HSV color space model; 3. The method for identifying the missing seedling quantity and missing seedling position of seedlings according to claim 1, characterized in that, The step of determining the root position of the seedlings according to the binary image of the seedling image specifically includes: Extracting the binary image of the seedling skeleton according to the binary image of the seedling image; Removing redundant branches in the binary image of the seedling skeleton to obtain a seedling morphology image; [[ID= 4. A recognition system for the number and position of missing seedlings of seedlings, which is applied to the recognition method for the number and position of missing seedlings of seedlings according to any one of claims 1-3, and is characterized in that, ​ ​ ​ A segmentation module, which is used to segment the HSV color space model by using an empirical threshold method to obtain a binary image of the seedling image; A root position determination module, which is used to determine the root position of the seedling according to the binary image of the seedling image; A fitting module, which is used to perform linear regression fitting by using the root position of the seedling to obtain a fitting straight line of the seedling; A calculation module, which is used to calculate the number of missing seedlings of the seedling according to the fitting straight line of the seedling; A coordinate determination module, which is used to determine the coordinates of the missing seedling position on the fitting straight line of the seedling according to the number of missing seedlings of the seedling.

5. The recognition system for the missing seedling quantity and missing seedling position of seedlings according to claim 4, characterized in that, The preprocessing module includes: A stitching sub-module, which is used to perform image stitching on the image information; A correction sub-module, which is used to perform orthophoto correction and multispectral image spectral correction on the stitched image information to obtain an RGB orthophoto image of the missing seedling detection target area; A conversion sub-module, which is used to convert the RGB orthophoto image into an HSV color space model.

6. The recognition system for the missing seedling quantity and missing seedling position of the seedlings according to claim 4, characterized in that The root position determination module includes: An extraction sub-module, which is used to extract a binary image of the seedling skeleton according to the binary image of the seedling image; A removal sub-module, which is used to remove redundant branches in the binary image of the seedling skeleton to obtain a seedling morphology image; A detection sub-module, which is used to perform corner detection on the seedling morphology image by using the Harris corner detection algorithm to determine the root position of the seedling.

7. The recognition system for the missing seedling quantity and missing seedling position of the seedlings according to claim 4, characterized in that, The calculation module includes: A projection point coordinate calculation sub-module, which is used to calculate the projection point coordinates of the root position of the seedling on the corresponding fitting straight line of the seedling; A spacing calculation sub-module, which is used to calculate the spacing between adjacent seedling projections in each row in the missing seedling detection target area, the spacing between the actual starting position and the set starting position of the seedlings in each row in the missing seedling detection target area, and the average spacing of the seedling projections according to the projection point coordinates; A first comparison sub-module, which is used to determine that there is a middle missing seedling between adjacent seedlings when the spacing between adjacent seedling projections is greater than a set distance; the set distance is a set multiple of the average spacing of the seedling projections; A second comparison sub-module, which is used to determine that there is a starting missing seedling between the actual starting position and the set starting position of the seedling when the spacing between the actual starting position and the set starting position of the seedling is greater than the set distance; A missing seedling number determination sub-module, which is used to calculate the number of missing seedlings at the middle missing seedling position and the number of missing seedlings at the starting missing seedling position according to the average spacing of the seedling projections to obtain the number of missing seedlings of the seedling.

8. The recognition system for the missing seedling quantity and missing seedling position of the seedlings according to claim 7, characterized in that, The coordinate determination module includes: A missing seedling projection point coordinate determination sub-module, which is used to determine the projection point coordinates of the root position of the missing seedling of the seedling on the corresponding fitting straight line of the seedling according to the number of missing seedlings of the seedling and the average spacing of the seedling projections; A missing seedling coordinate determination sub-module, which is used to determine the coordinates of the seedling at the missing seedling position on the fitting straight line of the seedling according to the projection point coordinates.

Images