CN116823918B - Crop seedling number measuring method, device, electronic equipment and storage medium - Google Patents
Crop seedling number measuring method, device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116823918B CN116823918B CN202311104512.8A CN202311104512A CN116823918B CN 116823918 B CN116823918 B CN 116823918B CN 202311104512 A CN202311104512 A CN 202311104512A CN 116823918 B CN116823918 B CN 116823918B
- Authority
- CN
- China
- Prior art keywords
- seedling
- planting area
- crop planting
- cell
- crop
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 77
- 238000003860 storage Methods 0.000 title claims abstract description 17
- 241000196324 Embryophyta Species 0.000 claims description 152
- 238000013507 mapping Methods 0.000 claims description 48
- 238000012795 verification Methods 0.000 claims description 31
- 238000005192 partition Methods 0.000 claims description 26
- 238000012545 processing Methods 0.000 claims description 25
- 239000002689 soil Substances 0.000 claims description 23
- 238000007476 Maximum Likelihood Methods 0.000 claims description 19
- 238000004590 computer program Methods 0.000 claims description 13
- 238000000691 measurement method Methods 0.000 claims description 4
- 238000000638 solvent extraction Methods 0.000 claims 3
- 238000012360 testing method Methods 0.000 abstract description 42
- 238000000605 extraction Methods 0.000 abstract description 26
- 238000009331 sowing Methods 0.000 abstract description 11
- 244000068988 Glycine max Species 0.000 description 23
- 235000010469 Glycine max Nutrition 0.000 description 22
- 238000010586 diagram Methods 0.000 description 17
- 230000006870 function Effects 0.000 description 10
- 238000012549 training Methods 0.000 description 9
- 238000009395 breeding Methods 0.000 description 7
- 230000001488 breeding effect Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 244000038559 crop plants Species 0.000 description 5
- 238000012937 correction Methods 0.000 description 4
- 238000011835 investigation Methods 0.000 description 4
- 239000000463 material Substances 0.000 description 4
- 238000010899 nucleation Methods 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 244000025254 Cannabis sativa Species 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 241000209140 Triticum Species 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 240000008042 Zea mays Species 0.000 description 1
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 1
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 235000005822 corn Nutrition 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000013095 identification testing Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000001568 sexual effect Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Processing (AREA)
Abstract
The invention provides a method, a device, electronic equipment and a storage medium for measuring the plant number of crop seedlings, belonging to the technical field of intelligent agriculture, wherein the method comprises the following steps: splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image; extracting seedling pixels from the crop planting area image, and determining the pixel quantity of each row of seedling in each cell in the crop planting area; and determining the number of seedling lines in each cell based on the number of pixels of each line of seedling lines in each cell. The invention can accurately calculate the number of the seedlings in each row of each cell, thereby realizing unmanned aerial vehicle high-throughput extraction of the number of the crop seedlings, being applicable to not only simple scenes with regular seedlings distribution under the condition of mechanical sowing, but also complex scenes such as uneven seedlings distribution, uneven row direction, stacking or ridge breaking and the like commonly existing in the existing artificial sowing, and greatly improving the efficiency and accuracy of extracting the number of the crop seedlings in each row of a large number of test cells.
Description
Technical Field
The invention relates to the technical field of intelligent agriculture, in particular to a method and a device for measuring the plant number of crop seedlings, electronic equipment and a storage medium.
Background
The seedling number is a crop early development index which is important in the identification of crop germplasm resources and the development of cell tests by variety breeding. The quantity of the seedlings not only reflects the quality and activity of crop seeds, but also is key basic information for crop growth management and yield potential evaluation.
Unlike the common field test, the crop resource identification and breeding test generally requires thousands or even tens of thousands of test cells to screen for dominant varieties and explore gene-environment interactions. The traditional investigation mode relying on manual seedling counting is faced with the problems of insufficient manpower, long time consumption and the like at present, measurement of all thousands of cells is difficult to complete in a short time, and investigation results are easy to record errors. Therefore, an efficient means for rapidly collecting the seedlings of a large number of test cells is urgently needed.
Disclosure of Invention
The invention provides a method, a device, electronic equipment and a storage medium for measuring the plant number of crop seedlings, which are used for solving the problems that the manual seedling counting investigation mode is insufficient in manpower, long in time consumption and the like in the prior art, the measurement of all thousands of cells is difficult to complete in a short time, and the investigation result is easy to record the fault condition.
The invention provides a method for measuring the plant number of crop seedlings, which comprises the following steps:
splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image;
extracting seedling pixels from the crop planting area image, and determining the pixel quantity of each row of seedling in each cell in the crop planting area;
and determining the number of seedling strains in each row in each cell based on the number of pixels of seedling strains in each row in each cell.
According to the method for measuring the plant number of the crop seedlings, provided by the invention, the seedling pixel extraction is carried out on the image of the crop planting area, and the pixel number of each row of seedling in each cell in the crop planting area is determined, and the method comprises the following steps:
performing pixel classification and image binarization processing on the crop planting area image to determine a binarization image of a seedling plant in the crop planting area;
and carrying out pixel partition statistics on the binarized image of the seedling in the crop planting area to obtain the pixel number of each row of seedling in each cell.
According to the method for measuring the plant number of the crop seedlings, which is provided by the invention, the pixel classification and the image binarization processing are carried out on the image of the crop planting area, and the binarization image of the seedling in the crop planting area is determined, and the method comprises the following steps:
Performing pixel classification on the crop planting area image by using a maximum likelihood classification method, and determining the object category to which each pixel in the crop planting area image belongs; the object categories include soil, weeds, seedlings and seedling shadows;
and setting all pixels of other object types except the seedling plants to null values in the crop planting area image, and generating a binarized image of the seedling plants in the planting area.
According to the method for measuring the plant number of the crop seedlings, provided by the invention, the pixel classification is carried out on the crop planting area image by using the maximum likelihood classification method, and the object category of each pixel in the crop planting area image is determined, which comprises the following steps:
acquiring verification point data according to sample points of various objects in the crop planting area image;
generating category characteristic data of the various objects according to the verification point data and the crop planting area image;
determining the attribution probability of each pixel in the crop planting area image to each object by using the category characteristic data of each object;
and determining the object category to which each pixel in the crop planting area image belongs according to the attribution probability of each pixel in the crop planting area image for various objects by taking the maximum attribution probability as a classification basis.
According to the method for measuring the plant number of the crop seedlings, provided by the invention, the pixel partition statistics is carried out on the binary image of the seedlings in the crop planting area to obtain the pixel number of each row of seedlings in each cell, and the method comprises the following steps:
obtaining geographic vector data of each row of planting areas in each cell;
and inputting the geographical vector data of each row of planting areas in each cell and the binarized images of the seedling plants in the planting areas into a preset partition statistical tool to obtain the pixel number of each row of seedling plants in each cell output by the preset partition statistical tool.
According to the method for measuring the plant number of the crop seedlings, which is provided by the invention, the number of the seedlings in each row in each cell is determined based on the pixel number of the seedlings in each row in each cell, and the method comprises the following steps:
substituting the pixel quantity of each row of seedling plants in each cell into a preset mapping relation to obtain the quantity of each row of seedling plants in each cell;
the preset mapping relation is used for representing the relation between the number of pixels of the crop seedlings and the number of plants of the crop seedlings; the preset mapping relation is obtained by performing data fitting according to the image samples of the crop planting area and the corresponding real seedling number.
According to the method for measuring the plant number of the crop seedlings, which is provided by the invention, the number of the seedlings in each row in each cell is determined based on the pixel number of the seedlings in each row in each cell, and the method comprises the following steps:
determining individual sizes of seedling strains in each cell;
determining a preset mapping relation applicable to the seedling in each cell according to the individual size of the seedling in each cell;
substituting the pixel quantity of each row of seedling plants in each cell into a corresponding applicable preset mapping relation to obtain the quantity of each row of seedling plants in each cell.
The invention also provides a device for measuring the plant number of the crop seedlings, which comprises:
the first splicing module is used for splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image;
the first processing module is used for extracting seedling pixels from the crop planting area image and determining the pixel quantity of each row of seedling in each cell in the crop planting area;
and the second processing module is used for determining the number of seedling strains in each row in each cell based on the number of pixels of seedling strains in each row in each cell.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the method for measuring the plant number of the crop seedlings when executing the program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements a method of measuring plant number of a crop plant as described in any one of the above.
The invention also provides a computer program product comprising a computer program which when executed by a processor implements a method of measuring plant number of a crop plant as described in any one of the above.
According to the crop seedling number measuring method, the device, the electronic equipment and the storage medium, the high-fraction code images of all frames of the crop planting area acquired by the unmanned aerial vehicle are spliced, so that a complete crop planting area image covering all large-batch test cells can be obtained, each object in the crop planting area image can be accurately classified and separated by utilizing a preset seedling pixel extraction mode, the image only containing crop seedlings in the crop planting area is determined, the pixel number of each row of seedlings in each cell is extracted, finally, the number of each row of seedlings in each cell is accurately calculated according to the pixel number of each row of seedlings in each cell by utilizing the direct connection between the seedling number and the seedling number, and therefore, unmanned aerial vehicle high-throughput extraction of the crop seedling number is realized, the method is not only suitable for a simple scene with relatively regular seedling distribution under the condition of machine sowing, but also suitable for a complex scene with uneven seedling distribution, uneven row direction, stacking or breaking existing in a crop planting resource identification and a variety test field, and the like, and the accuracy of the seedling number of each row of seedlings is greatly improved.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a method for measuring the number of plants in a crop plant;
fig. 2 is a schematic diagram of a mapping relationship between soybean pixel number and seedling number in the method for measuring crop seedling number according to the present invention;
FIG. 3a is a schematic diagram of a verification result obtained by the training set according to the power function mapping relation shown in FIG. 2;
FIG. 3b is a schematic diagram of a verification result obtained from the power function mapping relationship shown in FIG. 2 for a verification set provided by the present invention;
fig. 4 is a second schematic diagram of a mapping relationship between a number of soybean pixels and a number of seedlings in the method for measuring a number of seedlings of a crop according to the present invention;
FIG. 5a is a schematic diagram of a verification result obtained by the training set according to the mapping relation shown in FIG. 4;
FIG. 5b is a schematic diagram of a verification result obtained by the verification set according to the mapping formula shown in FIG. 4;
FIG. 6 is a second flow chart of the method for measuring plant number of crop seedlings according to the present invention;
FIG. 7 is a schematic structural view of a plant number measuring device for crop seedlings provided by the invention;
fig. 8 is a schematic diagram of the physical structure of the electronic device provided by the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The method, apparatus, electronic device and storage medium for measuring the plant number of the crop seedlings of the present invention are described below with reference to fig. 1 to 8.
Fig. 1 is a schematic flow chart of a method for measuring plant number of crop seedlings, as shown in fig. 1, including: step 110, step 120 and step 130.
Step 110, splicing all frame images of a crop planting area acquired by an unmanned aerial vehicle to obtain a crop planting area image;
Step 120, seedling pixel extraction is carried out on the crop planting area image, and the pixel number of each row of seedling in each cell in the crop planting area is determined;
and 130, determining the number of seedling lines in each cell based on the number of pixels of each line of seedling lines in each cell.
Specifically, the crop planting area described in the embodiment of the invention refers to the range of a whole resource variety breeding test area needing to extract the number of seedlings, and the test area consists of thousands of test cells planted with different variety materials of the same crop for phenotypic character observation, comparison and screening. The same crop can be soybean, wheat, corn or other crops.
The crop planting area image described in the embodiment of the invention refers to a crop planting area image which is obtained by splicing all high-fraction code images obtained by automatic continuous flight shooting of an unmanned aerial vehicle along a route and completely covers all test cells in the crop planting area.
Unmanned aerial vehicle remote sensing is a new means for acquiring field crop phenotype information by virtue of the advantages of high space-time resolution, good real-time performance, rapidness, low cost and the like. Especially in recent years, with the development of light and small sensors and the enhancement of the bearing capacity of platforms, unmanned aerial vehicle remote sensing technology is paid attention to and applied to crop resource identification and breeding research.
In the embodiment of the present invention, in step 110, an unmanned aerial vehicle carrying a high-fraction code sensor performs image acquisition on all cells of a crop planting test area. For the scene of soybean seedling stage (2 weeks after sowing, 12-16 days, growth stage: two-leaf stage V2 to three-leaf stage V3), an unmanned plane carrying a digital sensor with 2000 ten thousand pixels can be utilized to automatically and continuously fly along a route, one Zhang Degao fraction code image of a test area is shot, the flying height can be set to 12m (the resolution is 3 mm), and the heading overlapping rate and the side overlapping rate can be set to 80%. Meanwhile, a Real-Time Kinematic (RTK) thousands-looking-cm positioning system is utilized to collect the longitude and latitude information of the geographic position of the ground control point of the test area.
Further, in step 110, all the high-score digital images of the collected crop planting area are spliced by using unmanned aerial vehicle data processing software, so as to obtain a complete orthographic image of the crop planting area. Meanwhile, the geographical registration function of Arcgis software and high-precision geographical position information of RTK ground control points can be utilized to carry out geographical correction on the orthographic images of the crop planting areas, so that a complete image of the test area with correct spatial positions is obtained, and the crop planting area images are obtained.
In the embodiment of the present invention, in step 120, the influence of the background such as soil, weeds, seedling shadows and the like can be removed by using the feature extraction tool and the maximum likelihood classification tool of Arcgis software, and the seedling pixels can be extracted from each cell in the crop planting area image, so as to obtain the pixel number of each row of seedling in each cell.
Based on the foregoing embodiments, as an optional embodiment, step 120 of performing seedling pixel extraction on the crop planting area image, determining the number of pixels of each row of seedling in each cell in the crop planting area includes:
performing pixel classification and image binarization processing on the crop planting area image to determine a binarization image of a seedling plant in the crop planting area;
and carrying out pixel partition statistics on the binarized image of the seedling in the crop planting area to obtain the pixel quantity of the seedling in each row of each cell.
Specifically, in the embodiment of the invention, pixel classification processing can be performed on the crop planting area image, other objects affecting seedling pixel extraction in the crop planting area image are classified and removed, namely binarization processing is performed on the crop planting area image, so that binarization images of seedlings and backgrounds in the crop planting area are obtained.
Based on the foregoing embodiments, as an optional embodiment, performing pixel classification and image binarization processing on the crop planting area image to determine a binarized image of the seedling in the crop planting area, including:
carrying out pixel classification on the crop planting area image by using a maximum likelihood classification method, and determining the object category to which each pixel in the crop planting area image belongs; object categories include soil, weeds, plantlets, and plantlet shadows;
in the crop planting area image, all pixels of other object categories except for the seedling plants are set to null values, and a binarized image of the seedling plants in the crop planting area is generated.
It should be noted that, the Maximum likelihood classification method (Maximum-likelihood Classification) refers to an image classification method for classifying by establishing a nonlinear discrimination function set by a statistical method according to a Maximum likelihood ratio bayesian discrimination criterion method in two or more classes of discrimination, assuming that various distribution functions are normally distributed, selecting a training area, and calculating the attribution probability of each sample area to be classified.
Specifically, in the embodiment of the invention, all pixels in the crop planting area image can be classified by using a maximum likelihood classification method, each pixel in the crop planting area image is classified into the category with the highest attribution probability according to the attribution probability of each pixel in the discrimination image for each category, so that the object category of each pixel in the crop planting area image is determined, and the objects in the crop planting area image are classified into four categories of soil, weed, seedling and seedling shadow.
Optionally, in some embodiments, performing pixel classification on the crop planting area image by using a maximum likelihood classification method, determining an object class to which each pixel in the crop planting area image belongs includes:
acquiring verification point data according to sample points of various objects in the crop planting area image;
generating category characteristic data of various objects according to the verification point data and the crop planting area image;
determining the attribution probability of each pixel in the crop planting area image for various objects by utilizing the category characteristic data of various objects;
and determining the object category to which each pixel in the crop planting area image belongs according to the attribution probability of each pixel in the crop planting area image for various objects by taking the maximum attribution probability as a classification basis.
Specifically, the sample points described in the embodiments of the present invention are obtained by randomly extracting a plurality of pixel points from each object class in the crop planting area image.
The verification point data described in the embodiments of the present invention refers to a pixel sample point for verifying the classification accuracy of an object in a graph, and a portion may be selected from the pixel sample point as the verification point data.
The category characteristic data described in the embodiment of the invention refer to characteristic vector data which is used for representing each object category in the crop planting area image.
Although the conventional spectral index or color index thresholding method can distinguish seedlings from soil background well, it is not good to distinguish weeds, especially in the case of more weeds.
In an embodiment of the present invention, crop planting area images are classified into four types of objects of soil, weeds, seedlings and shadows using a maximum likelihood classification method. Specifically, an authentication point data file may be first created using an image processing tool in Arcgis software. A new shape file creating tool (Create New Shapefile) in Arcgis is utilized to create a point vector file (. Shp), a class field is added to the vector file, the class field comprises four classes of soil (soil), weeds (weeds), seedlings (soybean) and seedlings shadows (shadow), then a plurality of sample points are randomly selected for each class (soil, weeds, seedlings and shadows) on a crop planting area image, wherein the soil, shadows and seedlings are easy to distinguish, tens of sample points are needed, the separation difficulty of the weeds and the seedlings is large, and the number of sample points of the two types is more. Further, the selected sample point data may be used as a verification point data file.
And secondly, creating a feature file, namely creating category feature data of various objects. The output profile may be created based on the verification point data file and the crop planting area image using a create profile (Create Signatures) tool under multivariate (multisariate) in Arcgis, thereby generating category profile data for various types of objects.
Further, by using a maximum likelihood classification tool (Maximum Likelihood Classification) in Arcgis, classification processing is performed on the crop planting area image based on the generated category characteristic data file, and the output storage position and the name after image classification are set, and other settings are generally default. And judging the attribution probability of each pixel in the crop planting area image for each category according to a statistical method and a category characteristic data file through processing of a maximum likelihood classification method tool, and classifying the pixel into the category with the maximum attribution probability by taking the maximum attribution probability as a classification basis, so that the soil, the weeds, the seedling and the shadow objects in the crop planting area image can be identified and distinguished. By default, all pixels in the crop planting area image are classified, and the weights (priori probability weighting) of the class prior probabilities are equal.
According to the method provided by the embodiment of the invention, the verification point data of various objects in the crop planting area image is obtained, the category characteristic data of various objects is extracted, each pixel in the crop planting area image can be classified according to the category characteristic data of various objects by means of the maximum likelihood classification tool in Arcgis, the object category to which each pixel in the crop planting area image belongs is accurately classified, the classification precision of the background object in the crop planting area image is improved, and the method is suitable for complex scenes with more background objects and high weed separation difficulty.
Further, in the embodiment of the present invention, according to the classified crop planting area image, all pixel values belonging to the seedling plant may be set to 1 in the crop planting area image, and all pixel values of other object classes except the seedling plant may be set to null, so as to obtain a binarized image of the crop planting area with only the seedling plant and the non-seedling plant background.
In one embodiment, based on the obtained image of the crop planting area classified into soil, weeds, seedlings and shadows, seedling class pixels in the image are extracted individually by using an attribute extraction tool (Extract by attributes) of Arcgis, and other class pixels are all null, so that a binary image of seedling and background is obtained. Meanwhile, soil, weeds and shadows in the four classified images can be set into one type by using a classification tool (Reclassify) of Arcgis, and seedlings are reserved, so that a binary image of seedlings and non-seedlings background is obtained.
According to the method provided by the embodiment of the invention, the maximum likelihood classification method is utilized to classify the pixels of the crop planting area image, and all the pixels in the crop planting area image are classified into soil, weeds, seedling plants and seedling plant shadow categories, so that the image binarization processing of independently extracting seedling plant pixels can be performed, the crop planting area binarization image of the pure seedling plant is obtained, the operation method is simple and effective, the extraction effect of the seedling plant pixels is greatly improved, and the accuracy and reliability of the number of the subsequently extracted seedling plants are improved.
Based on the foregoing embodiment, as an optional embodiment, performing pixel partition statistics on the binary image of the seedling in the crop planting area to obtain the number of pixels of each row of seedling in each cell, where the method includes:
obtaining geographic vector data of each row of planting areas in each cell;
and (3) inputting the geographic vector data of each row of planting areas in each cell and the binarized images of the seedling plants in the planting areas into a preset partition statistical tool to obtain the pixel number of each row of seedling plants in each cell output by the preset partition statistical tool.
Specifically, the geographic vector data described in the embodiment of the invention refers to vector shp file data which is generated by digitizing a cell layout of a crop planting test area by Arcgis software and contains all rows of cells, name numbers and planting material information (seed variety and the like).
The preset partition statistics tool described in embodiments of the present invention may employ partition statistics output of Arcgis software to a table tool (Zonal Statistics as Table).
Further, after the digitized vector diagram of the test area presenting the geographical position of each cell is obtained, the geographical vector data and the binary image after only pure seedlings are extracted after the soil, weeds and shadows are removed from the crop planting area can be Input into a preset partition statistics tool, specifically, partition statistics in Arcgis software can be utilized to output to a table tool (Zonal Statistics as Table), and a geographical vector data shp file of the test cell layout diagram is Input through a zone Input module (Input zone data) so as to clearly determine zone information according to the operation from the partition statistics to the table.
Then, using an input grid (Input value raster) module, a crop planting area binarized image of the pure seedlings after the shading of the soil, the weeds and the seedlings is removed is input, so that the object for carrying out the regional statistics to the table operation is definitely confirmed. And setting the name and the storage position of a statistical table of a partition statistical-to-table operation result Output in a data Output module (Output table), and then executing the partition statistical-to-table operation to obtain a statistical table of the pixel numbers of all rows of seedlings of all cells, and obtaining the pixel numbers of all rows of seedlings in each cell.
In the embodiment of the invention, the partition statistics in Arcgis software is utilized to output to the surface tool, and the pixel partition statistics is carried out on the binary image of the crop planting area after the seedling plants are removed from soil, weeds and shadows, so that the pixel number of each row of seedling plants in each cell can be effectively extracted, the extraction precision is high, and the subsequent seedling number extraction precision is further improved.
According to the method, the image of the crop planting area is subjected to pixel classification and image binarization processing by utilizing Arcgis software, so that a binarization image of a pure seedling plant is obtained, further, pixel partition statistics is carried out on the binarization image of the pure seedling plant of the crop planting area, the pixel number of the seedling plant in each row of each cell is extracted, the method is easy to operate, the extraction precision of the pixel number of the seedling plant is high, and the method is not only suitable for simple scenes with regular seedling plant distribution under the mechanical sowing condition, but also suitable for complex scenes with uneven seedling plant distribution, uneven row direction, stacked or broken ridges, larger grass damage and the like which are commonly existed in crop resource identification and variety area test fields by adopting manual sowing.
Further, in step 130, the number of seedling plants is directly calculated according to the number of pixels of each row of seedling plants in each cell by utilizing the relation between the number of seedling plants and the number of pixels of the seedling plants, and meanwhile, the influence of individual size differences of the seedling plants is also considered, so that the unmanned aerial vehicle high-throughput extraction of the number of seedling plants is realized, and the accuracy can reach about 0.9.
According to the crop seedling number measuring method, the high-fraction code images of one frame of the crop planting area collected by the unmanned aerial vehicle are spliced, a crop planting area image which completely covers all large-batch test cells can be obtained, the soil, weeds, seedlings and various objects of shadows of the seedlings in the crop planting area image can be accurately classified and separated by utilizing a preset seedling pixel extraction mode, so that a crop planting area image only containing crop seedlings is obtained, the pixel number of each row of seedlings in each cell is extracted, finally, the number of each row of seedlings in each cell is accurately calculated according to the pixel number of each row of seedlings in each cell by utilizing the relation between the number of seedlings and the pixel number of each seedling, and therefore the unmanned aerial vehicle high-throughput extraction of the crop seedlings is realized, the unmanned aerial vehicle is not only suitable for a simple scene with relatively regular seedling distribution under the condition of machine sowing, but also suitable for a scene with uneven distribution, uneven row direction, collection or broken weed, large-pile and the like of the seedlings in each row of each cell test field by adopting artificial seeding, and the test field is greatly improved in the accuracy of the test field of each row of the crop seedlings and the test field.
Based on the content of the above embodiments, as an alternative embodiment, determining the number of seedlings of each row of each cell based on the number of pixels of each row of seedlings of each cell includes:
substituting the pixel quantity of each row of seedling plants of each cell into a preset mapping relation to obtain the quantity of each row of seedling plants of each cell;
the preset mapping relation is used for representing the relation between the number of pixels of the crop seedlings and the number of plants of the crop seedlings; the preset mapping relation is obtained by carrying out data fitting according to the image samples of the crop planting areas and the corresponding real seedling numbers.
Specifically, the preset mapping relation described in the embodiment of the present invention refers to a mapping relation obtained by performing data statistics and data fitting in advance according to the number of pixels of the crop seedling and the number of the corresponding real seedlings of the image sample of the crop planting area, and can well represent the mapping relation between the number of pixels of the crop seedling and the number of the crop seedlings. The image sample of the crop planting area may be image data of a part of cells (rows) selected from the crop planting area image collected by the unmanned aerial vehicle.
In a specific embodiment of the present invention, image samples of 312 rows of the test cell and the corresponding number of true seedlings are obtained, so as to obtain a sample set, and the sample set may be divided into a training set and a verification set according to a preset ratio, for example, generally 2:1. And carrying out data statistics and data fitting by using the training set to obtain a mapping relation, and evaluating the accuracy of the mapping relation by using the verification set.
FIG. 2 is a schematic diagram showing the mapping relationship between the number of soybean pixels and the number of seedlings in the method for measuring the number of seedlings in crops, wherein the ordinate represents the number of seedlings in crops, the abscissa represents the number of seedlings in crops, the sample specification of data points is 1 row, the length is 2m (1 row×2 m), and the method is specifically based on the flight height of an unmanned aerial vehicle of 12m, and high-fraction digital image data collected 12-16 days (growth period V2-V3) after soybean sowing are respectively processedAnd (5) performing sexual fitting and power function fitting. Wherein the certainty factor R of the linear fit 2 For 0.5641, the deterministic coefficient R of the power function fit 2 0.6642.
Further, after the preset mapping relation is determined, the pixel number of each row of seedling plants in each cell can be substituted into the preset mapping relation to obtain the number of each row of seedling plants in each cell. Substituting the number of pixels of soybean seedlings in each row of each cell into R in FIG. 2 2 In the larger power function mapping relation, the number of soybean seedlings in each cell can be obtained.
Fig. 3a is a schematic diagram of a verification result obtained by the training set according to the power function mapping relation shown in fig. 2, and fig. 3b is a schematic diagram of a verification result obtained by the verification set according to the power function mapping relation shown in fig. 2. As shown in FIGS. 3a and 3b, the accuracy R of the training set and the validation set 2 0.5826 and 0.4336, respectively, which can reflect the relative trend of the seedling number of each cell as a whole, the specific numerical precision of the seedling number of each cell is not high enough.
Therefore, the invention further refines the mapping relation between the soybean pixel number and the seedling number according to the seedling size by considering the influence of the seedling size difference on the relation between the soybean pixel number and the seedling number, and accurately calculates the corresponding seedling number according to the refined mapping relation after obtaining the accurate and reliable seedling number, thereby realizing the high-precision and high-flux extraction of the crop seedling number and improving the precision of the extraction of the crop seedling number.
Based on the content of the foregoing embodiments, as an alternative embodiment, determining the number of seedling lines in each cell based on the number of pixels of each line in each cell includes:
determining individual sizes of seedling strains in each cell;
determining a preset mapping relation applicable to the seedling plants in each cell according to the individual sizes of the seedling plants in each cell;
substituting the pixel quantity of each row of seedling plants in each cell into a corresponding applicable preset mapping relation to obtain the quantity of each row of seedling plants in each cell.
Specifically, in the embodiment of the invention, the influence of the size difference of the seedlings can be further considered to obtain a more accurate mapping relation.
Fig. 4 is a second schematic diagram of a mapping relationship between soybean pixel numbers and seedling numbers in the crop seedling number measurement method provided by the invention, as shown in fig. 4, in an embodiment of the invention, an influence of a seedling size difference on a soybean pixel number-seedling number relationship is considered, wherein the seedling size can be divided into four types of sizes of small plants, medium and large plants, the small plants occupy about 50 to 150 pixels, the medium and small plants occupy about 150 to 200 pixels, the medium and large plants occupy about 200 to 260 pixels, and the large plants occupy about 260 to 350 pixels. The four fitting curves shown in fig. 4 are mapping relation formulas of small, medium and large plants, medium and small plants and large plants of seedling size from left to right, and are specifically obtained by linear fitting of high-fraction digital image data acquired 12-16 days (growth period V2-V3) after soybean sowing based on unmanned aerial vehicle flight height of 12m, wherein the mapping relation formulas of the small, medium and large plants, medium and small plants and the large plants correspond to the accuracy R 2 0.9397, 0.9118, 0.9135 and 0.9122, respectively.
In the embodiment of the invention, a place where seedlings are not connected and not stacked on the seedling binarization image is required to be selected, a plurality of samples which can be determined as single plants are checked by manual visual observation, and the sizes of the seedlings are investigated, so that the individual sizes of the seedlings in a crop planting area are determined, and the number of samples can be increased according to the requirement aiming at the fact that the individual differences of the seedlings in a test area are relatively large. Then, according to the individual sizes of the seedling plants in each region, an applicable mapping relation can be selected, and then the pixel number of the seedling plants in each row in each region is substituted into the corresponding mapping relation, so that the seedling number is calculated.
Fig. 5a is a schematic diagram of a verification result obtained by the training set according to the mapping relation shown in fig. 4, and fig. 5b is a schematic diagram of a verification result obtained by the verification set according to the mapping relation shown in fig. 4. As shown in fig. 5a and 5b, the seedling measurement method in the present embodiment is compared with the method in which the individual difference of seedlings is not consideredThe accuracy of the method is greatly improved, and the certainty coefficient R of the training set and the verification set 2 0.9209 and 0.8907 are respectively achieved, actual demands can be met, and a technical approach for rapidly extracting the seedling number of each row of each cell is provided for the crop resource identification of artificial seeding and the complex distribution scene of actual seedlings in the variety breeding test field.
According to the method provided by the embodiment of the invention, the mapping relation is constructed by utilizing the direct connection between the seedling number and combining the influence factors of the seedling size, after the accurate and reliable seedling number is obtained, the corresponding seedling number can be calculated more accurately through the mapping relation, so that the high-precision and high-flux extraction of the crop seedling number is realized, and meanwhile, the precision of the information extraction of the crop seedling number is greatly improved.
Fig. 6 is a second flow chart of the method for measuring the plant number of the crop seedlings, as shown in fig. 6, and the method specifically includes: collecting data; image stitching correction and test area digitization; separating the seedling from the background to generate a binarized image; batch extraction of cells (or each row) Miao Zhushu.
First, data acquisition is performed. Specifically, in the soybean seedling stage (2 weeks to 16 days after sowing, the growth period V2 to V3), unmanned aerial vehicle carrying high-score digital sensors is used for collecting high-score code images of soybean resource identification breeding test areas (namely crop planting areas), the flying height is 12m (the resolution is 3 mm), and the heading overlapping rate and the side overlapping rate are both 80%. And meanwhile, the RTK thousands of seeking-cm positioning system is utilized to collect the longitude and latitude information of the geographic position of the ground control point of the test area. Therefore, the unmanned aerial vehicle can automatically and continuously collect high-fraction code images of one frame of the test area along the aerial line above the crop planting area, and a high-fraction code image set shot in the current flight of the crop planting area is obtained.
Secondly, image stitching correction is carried out, and the test area is digitized. Specifically, the acquired frame number digital images can be spliced by using drawing software to obtain a complete test area orthophoto. And the geographic registration function of Arcgis software and the longitude and latitude data of RTK ground control points can be utilized to carry out geographic correction on the orthographic image of the test area, so that a complete image of the test area with correct spatial position is obtained, and a crop planting area image is obtained. Further, according to the soybean resource identification test (material) cell layout, the test cell layout is digitized by Arcgis software, and a vector shp file containing all lines of all cells of the test area and the planting material information thereof is generated for carrying out subsequent image partition statistics to extract the pixel number of each line of each cell.
Then, the seedlings are separated from the background, and a binarized image is generated. Specifically, arcgis software can be utilized to randomly select seedling plants, shadows, soil and weed sample points on the crop planting area image as verification files, namely verification point data files are obtained. Furthermore, a feature file creating tool in Arcgis software can be utilized to automatically generate various feature files by taking the verification file as input, so as to obtain category feature data files of various objects in the crop planting area image. And (3) separating seedling plants from the background such as shadow, soil, weeds and the like by using a maximum likelihood classification tool and taking a crop planting area image and a characteristic file as inputs, so as to obtain a crop planting area binary image of the pure seedling plants.
And finally, carrying out batch extraction of the seedling number of each row in each cell. Specifically, the partition statistics output by Arcgis software can be utilized to a table tool, and the binary image of the pure seedling plant and the digitized cell layout shp file of the test area are taken as input to output the seedling plant pixel count statistics table of all rows of all cells. Further, a plurality of individual plant samples are selected at the places where the seedling plants are discrete on the image, the size range of the individual seedling plants is inspected, and the number of soybean seedling plants is extracted based on the mapping relation between the number of soybean pixels and the number of soybean seedling plants taking the difference of the seedling plants into consideration, so that the number of soybean seedling plants in each row of each cell is obtained.
According to the method provided by the embodiment of the invention, the advantages of rapidness, real time, low cost and high space-time resolution of an unmanned aerial vehicle remote sensing technology in crop resource identification and variety breeding test observation are fully utilized, aiming at complex scenes that actual seedlings are not regularly distributed due to manual sowing, the row direction is not straight, stacking, ridge breaking, grass are multiple and the like, and the traditional means and the prior background technical scheme are difficult to adapt to, based on a soybean seedling period unmanned aerial vehicle high-fraction digital image, the direct connection between the plant number and the plant pixel number is utilized, and meanwhile, the individual size difference of the plant and the influence of weeds are considered, so that the unmanned aerial vehicle high-throughput extraction method for the soybean plant number is provided, and the precision is about 0.9.
The device for measuring the plant number of the crop seedlings provided by the invention is described below, and the device for measuring the plant number of the crop seedlings described below and the method for measuring the plant number of the crop seedlings described above can be correspondingly referred to each other.
Fig. 7 is a schematic structural diagram of a plant number measuring device for crop seedlings according to the present invention, as shown in fig. 7, including:
the first stitching module 710 is configured to stitch all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image;
the first processing module 720 is configured to perform seedling pixel extraction on the crop planting area image, and determine the number of pixels of each row of seedlings in each cell in the crop planting area;
and a second processing module 730, configured to determine the number of seedling lines in each cell based on the number of pixels of each line in each cell.
The crop seedling number measuring device in this embodiment may be used to execute the embodiment of the crop seedling number measuring method, and its principle and technical effects are similar, and will not be described herein again.
According to the crop seedling number measuring device, the images of the crop planting areas acquired by the unmanned aerial vehicle are spliced in a frame, so that a complete crop planting area image covering a large number of test cells can be obtained, soil, weeds, seedlings and shadows in the crop planting area image can be accurately classified and separated by utilizing a preset seedling pixel extraction mode, the image of the crop seedling only contained in the crop planting area is determined, the pixel number of each row of seedlings in each cell is extracted, finally, the number of each cell and each row of seedlings is accurately calculated according to the pixel number of each row of seedlings in each cell by utilizing the relation between the number of seedlings and the pixel number of each seedling, and therefore the unmanned aerial vehicle high-throughput extraction of the crop seedling number is realized, the unmanned aerial vehicle seedling number measuring device is suitable for a simple scene with relatively regular seedling distribution under the condition of machine seeding, and is also suitable for complex scenes such as uneven seedling distribution, stacking or ridge cutting, more weeds and the like which are commonly existed in artificial seeding, and the crop resource identification and seed selection test are adopted, and the accuracy of the crop seedling number of each row of seedlings is greatly improved.
Fig. 8 is a schematic diagram of an entity structure of an electronic device according to the present invention, as shown in fig. 8, the electronic device may include: processor 810, communication interface (Communications Interface) 820, memory 830, and communication bus 840, wherein processor 810, communication interface 820, memory 830 accomplish communication with each other through communication bus 840. The processor 810 may invoke logic instructions in the memory 830 to perform the method for measuring the number of seedlings provided by the methods described above, the method comprising: splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image; extracting seedling pixels from the crop planting area image, and determining the pixel quantity of each row of seedling in each cell in the crop planting area; and determining the number of seedling strains in each row in each cell based on the number of pixels of seedling strains in each row in each cell.
Further, the logic instructions in the memory 830 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product, the computer program product comprising a computer program, the computer program being storable on a non-transitory computer readable storage medium, the computer program, when executed by a processor, being capable of executing the method for measuring plant number of crop plants provided by the above methods, the method comprising: splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image; extracting seedling pixels from the crop planting area image, and determining the pixel quantity of each row of seedling in each cell in the crop planting area; and determining the number of seedling strains in each row in each cell based on the number of pixels of seedling strains in each row in each cell.
In yet another aspect, the present invention provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the method for measuring plant number of crop plants provided by the above methods, the method comprising: splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image; extracting seedling pixels from the crop planting area image, and determining the pixel quantity of each row of seedling in each cell in the crop planting area; and determining the number of seedling strains in each row in each cell based on the number of pixels of seedling strains in each row in each cell.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (5)
1. A method for measuring the number of plants in a crop, comprising:
splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image;
extracting seedling pixels from the crop planting area image, and determining the pixel quantity of each row of seedling in each cell in the crop planting area;
determining the number of seedling strains in each row in each cell based on the number of pixels of seedling strains in each row in each cell;
the step of extracting seedling pixels from the crop planting area image to determine the pixel number of each row of seedling in each cell in the crop planting area comprises the following steps:
Performing pixel classification and image binarization processing on the crop planting area image to determine a binarization image of a seedling plant in the crop planting area;
carrying out pixel partition statistics on the binarized image of the seedling in the crop planting area to obtain the pixel number of each row of seedling in each cell;
the pixel classification and image binarization processing are carried out on the crop planting area image, and the determination of the binarization image of the seedling plant in the crop planting area comprises the following steps:
performing pixel classification on the crop planting area image by using a maximum likelihood classification method, and determining the object category to which each pixel in the crop planting area image belongs; the object categories include soil, weeds, seedlings and seedling shadows;
in the crop planting area image, all pixels of other object categories except for the seedling plants are set to null values, and a binarized image of the seedling plants in the planting area is generated;
the method for classifying pixels of the crop planting area image by using the maximum likelihood classification method, determining the object category to which each pixel belongs in the crop planting area image, includes:
acquiring verification point data according to sample points of various objects in the crop planting area image;
Generating category characteristic data of the various objects according to the verification point data and the crop planting area image;
determining the attribution probability of each pixel in the crop planting area image to each object by using the category characteristic data of each object;
taking the maximum attribution probability as a classification basis, and determining the object category to which each pixel in the crop planting area image belongs according to the attribution probability of each pixel in the crop planting area image for various objects;
the pixel partition statistics is performed on the binarized image of the seedling in the crop planting area to obtain the pixel number of each row of seedling in each cell, including:
obtaining geographic vector data of each row of planting areas in each cell;
the geographical vector data of each row of planting areas in each cell and the binarized images of the seedling plants in the planting areas are input into a preset partition statistical tool to obtain the pixel number of each row of seedling plants in each cell output by the preset partition statistical tool;
the determining the number of seedling lines in each cell based on the number of pixels of each line of seedling lines in each cell comprises the following steps:
Substituting the pixel quantity of each row of seedling plants in each cell into a preset mapping relation to obtain the quantity of each row of seedling plants in each cell;
the preset mapping relation is used for representing the relation between the number of pixels of the crop seedlings and the number of plants of the crop seedlings; the preset mapping relation is obtained by performing data fitting according to the image samples of the crop planting area and the corresponding real seedling number.
2. The method of claim 1, wherein determining the number of seedlings in each row in each cell based on the number of pixels in each row in each cell, comprises:
determining individual sizes of seedling strains in each cell;
determining a preset mapping relation applicable to the seedling in each cell according to the individual size of the seedling in each cell;
substituting the pixel quantity of each row of seedling plants in each cell into a corresponding applicable preset mapping relation to obtain the quantity of each row of seedling plants in each cell.
3. A plant number measuring device for crop seedlings, comprising:
the first splicing module is used for splicing all frame images of the crop planting area acquired by the unmanned aerial vehicle to obtain a crop planting area image;
The first processing module is used for extracting seedling pixels from the crop planting area image and determining the pixel quantity of each row of seedling in each cell in the crop planting area;
the second processing module is used for determining the number of seedling strains in each row in each cell based on the number of pixels of seedling strains in each row in each cell;
wherein the first processing module comprises:
the first processing submodule is used for carrying out pixel classification and image binarization processing on the crop planting area image and determining a binarization image of the seedling plant in the crop planting area;
the first partitioning sub-module is used for carrying out pixel partitioning statistics on the binarized image of the seedling in the crop planting area to obtain the pixel number of each row of seedling in each cell;
the first processing sub-module is specifically configured to:
performing pixel classification on the crop planting area image by using a maximum likelihood classification method, and determining the object category to which each pixel in the crop planting area image belongs; the object categories include soil, weeds, seedlings and seedling shadows;
in the crop planting area image, all pixels of other object categories except for the seedling plants are set to null values, and a binarized image of the seedling plants in the planting area is generated;
The first processing sub-module is specifically configured to:
acquiring verification point data according to sample points of various objects in the crop planting area image;
generating category characteristic data of the various objects according to the verification point data and the crop planting area image;
determining the attribution probability of each pixel in the crop planting area image to each object by using the category characteristic data of each object;
taking the maximum attribution probability as a classification basis, and determining the object category to which each pixel in the crop planting area image belongs according to the attribution probability of each pixel in the crop planting area image for various objects;
the first partitioning submodule is specifically configured to:
obtaining geographic vector data of each row of planting areas in each cell;
the geographical vector data of each row of planting areas in each cell and the binarized images of the seedling plants in the planting areas are input into a preset partition statistical tool to obtain the pixel number of each row of seedling plants in each cell output by the preset partition statistical tool;
the second processing module is specifically configured to:
substituting the pixel quantity of each row of seedling plants in each cell into a preset mapping relation to obtain the quantity of each row of seedling plants in each cell;
The preset mapping relation is used for representing the relation between the number of pixels of the crop seedlings and the number of plants of the crop seedlings; the preset mapping relation is obtained by performing data fitting according to the image samples of the crop planting area and the corresponding real seedling number.
4. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the crop seedling number measurement method of any one of claims 1 to 2 when the program is executed.
5. A non-transitory computer readable storage medium having stored thereon a computer program, characterized in that the computer program, when executed by a processor, implements the crop seedling number measurement method according to any one of claims 1 to 2.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311104512.8A CN116823918B (en) | 2023-08-30 | 2023-08-30 | Crop seedling number measuring method, device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311104512.8A CN116823918B (en) | 2023-08-30 | 2023-08-30 | Crop seedling number measuring method, device, electronic equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116823918A CN116823918A (en) | 2023-09-29 |
CN116823918B true CN116823918B (en) | 2023-12-22 |
Family
ID=88127783
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311104512.8A Active CN116823918B (en) | 2023-08-30 | 2023-08-30 | Crop seedling number measuring method, device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116823918B (en) |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102339378A (en) * | 2010-07-22 | 2012-02-01 | 中国农业机械化科学研究院 | Method and device for automatically extracting cotton seeds |
CN108985260A (en) * | 2018-08-06 | 2018-12-11 | 航天恒星科技有限公司 | A kind of remote sensing and meteorological integrated rice yield estimation method |
CN112614147A (en) * | 2020-12-24 | 2021-04-06 | 中国农业科学院作物科学研究所 | Method and system for estimating plant density of crop at seedling stage based on RGB image |
CN115358991A (en) * | 2022-08-19 | 2022-11-18 | 浙江大学华南工业技术研究院 | Method and system for identifying seedling leaking quantity and position of seedlings |
US11553636B1 (en) * | 2022-06-21 | 2023-01-17 | FarmWise Labs Inc. | Spacing-aware plant detection model for agricultural task control |
-
2023
- 2023-08-30 CN CN202311104512.8A patent/CN116823918B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102339378A (en) * | 2010-07-22 | 2012-02-01 | 中国农业机械化科学研究院 | Method and device for automatically extracting cotton seeds |
CN108985260A (en) * | 2018-08-06 | 2018-12-11 | 航天恒星科技有限公司 | A kind of remote sensing and meteorological integrated rice yield estimation method |
CN112614147A (en) * | 2020-12-24 | 2021-04-06 | 中国农业科学院作物科学研究所 | Method and system for estimating plant density of crop at seedling stage based on RGB image |
US11553636B1 (en) * | 2022-06-21 | 2023-01-17 | FarmWise Labs Inc. | Spacing-aware plant detection model for agricultural task control |
CN115358991A (en) * | 2022-08-19 | 2022-11-18 | 浙江大学华南工业技术研究院 | Method and system for identifying seedling leaking quantity and position of seedlings |
Also Published As
Publication number | Publication date |
---|---|
CN116823918A (en) | 2023-09-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Xu et al. | Aerial images and convolutional neural network for cotton bloom detection | |
US20230292647A1 (en) | System and Method for Crop Monitoring | |
Sumesh et al. | Integration of RGB-based vegetation index, crop surface model and object-based image analysis approach for sugarcane yield estimation using unmanned aerial vehicle | |
CN106971167B (en) | Crop growth analysis method and system based on unmanned aerial vehicle platform | |
CN106951836B (en) | crop coverage extraction method based on prior threshold optimization convolutional neural network | |
CN105740759B (en) | Semilate rice information decision tree classification approach based on feature extraction in multi-temporal data | |
CN109325431B (en) | Method and device for detecting vegetation coverage in feeding path of grassland grazing sheep | |
CN108140118A (en) | For identifying the method and information system of at least one plant plant being planted on arable land | |
David et al. | Vegetation indices and textures in object-based weed detection from UAV imagery | |
Xu et al. | Classification method of cultivated land based on UAV visible light remote sensing | |
CN113029971A (en) | Crop canopy nitrogen monitoring method and system | |
CN108537164B (en) | Method and device for monitoring germination rate of dibbling and sowing based on unmanned aerial vehicle remote sensing | |
CN113011221A (en) | Crop distribution information acquisition method and device and measurement system | |
CN117576195A (en) | Plant leaf morphology recognition method | |
CN112734083A (en) | Rice harvester path planning control system based on machine vision | |
CN113850132A (en) | Crop identification method, device, equipment and storage medium | |
CN117036926A (en) | Weed identification method integrating deep learning and image processing | |
CN113377062B (en) | Multifunctional early warning system with disease and pest damage and drought monitoring functions | |
CN113570273B (en) | Spatialization method and system for irrigation farmland statistical data | |
CN115619286B (en) | Method and system for evaluating quality of sample plot of breeding field district | |
CN116823918B (en) | Crop seedling number measuring method, device, electronic equipment and storage medium | |
Ormeci et al. | Identification of crop areas Using SPOT–5 data | |
CN113240340B (en) | Soybean planting area analysis method, device, equipment and medium based on fuzzy classification | |
CN116258844A (en) | Rapid and accurate identification method for phenotype character of cotton leaf | |
de Souza et al. | Parameters of the phytosociological survey to evaluate the abundance, distribution, and diversity of the weed community |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |