CN109344843B - Method and device for extracting rice seedling row line, computer equipment and storage medium - Google Patents
Method and device for extracting rice seedling row line, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109344843B CN109344843B CN201811043572.2A CN201811043572A CN109344843B CN 109344843 B CN109344843 B CN 109344843B CN 201811043572 A CN201811043572 A CN 201811043572A CN 109344843 B CN109344843 B CN 109344843B
- Authority
- CN
- China
- Prior art keywords
- seedling
- row
- rice
- coordinate
- field image
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/231—Hierarchical techniques, i.e. dividing or merging pattern sets so as to obtain a dendrogram
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
- G06V10/457—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by analysing connectivity, e.g. edge linking, connected component analysis or slices
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method for extracting rice seedling row lines, which comprises the following steps: acquiring a rice seedling field image; identifying the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image; clustering a plurality of seedling coordinates, and identifying the seedling row in current operation; and performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line. By applying the pre-trained deep convolutional neural network and the clustering algorithm, the interference of environmental factors in the rice field is better eliminated, the current rice seedling row line can be more accurately identified, and the automatic driving of the rice transplanter is guided.
Description
Technical Field
The invention relates to the field of image processing, in particular to a method and a device for extracting rice seedling row lines, computer equipment and a storage medium.
Background
The agricultural field navigation technology provides navigation information for the operation of modern agricultural production machinery, and is applied to various agricultural machinery equipment. The importance of agricultural field navigation is particularly evident in agricultural equipment used for seeding.
At present, the research of agricultural field navigation at home and abroad mainly focuses on the aspect of machine vision navigation. Mengqing et al constructs a Cg component irrelevant to illumination, selects a 2Cg-Cr-Cb characteristic factor to perform graying processing on an image, establishes a constraint model of a corn crop row linear equation according to the characteristics of crop rows in the image, and performs optimization solution on the crop row linear equation by using a particle swarm algorithm to obtain a navigation line. In order to adapt to the phenomenon of unobvious image contrast or supersaturation caused by various illumination conditions, Romeo and the like design a crop-background image segmentation system based on image histogram analysis, and the system discriminates the contrast and the saturation of an image through a histogram to select a proper wheat seedling image segmentation method.
The accuracy of accurate identification of seedling belt image information and leading line extraction is the premise of realizing accurate operation of sowing equipment. In the implementation process of identifying rice seedling row lines and navigating a rice transplanter by applying a traditional machine vision method, the inventor finds that at least the following problems exist: in the transplanting stage, the field environment of rice is complex, the water layer is thin, and the traditional image identification method is difficult to accurately identify the rice seedling row line and guide the rice transplanter to drive autonomously.
Disclosure of Invention
Based on this, it is necessary to provide a method, an apparatus, a computer device and a storage medium for extracting a row line of rice seedlings, aiming at the problem that the identification of the row line of rice seedlings is not accurate enough.
In one aspect, the embodiment of the invention provides a method for extracting rice seedling row lines, which comprises the following steps:
acquiring a rice seedling field image;
recognizing the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image;
clustering a plurality of seedling coordinates, and identifying the seedling row in current operation;
and performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line.
In one embodiment, before acquiring the field image of the rice seedling, the method further comprises:
acquiring sample data of a field image of the rice seedling, wherein the sample data of the field image of the rice seedling comprises field images of the rice seedling of various natural conditions, illumination angles and varieties;
and training the deep convolution neural network model according to the sample data of the rice seedling field image.
In one embodiment, the deep convolutional neural network model is a fast RCNN neural network model.
In one embodiment, clustering the seedling coordinates and identifying the currently operating seedling row comprises:
classifying the seedling coordinates through a clustering algorithm to obtain row classification information corresponding to each seedling coordinate;
and identifying the seedling row in the current operation according to each seedling coordinate, the row classification information corresponding to each seedling coordinate and the central coordinate of the field image of the rice seedling.
In one embodiment, the clustering algorithm is a hierarchical clustering algorithm.
In one embodiment, the process of identifying the currently operating seedling row according to each seedling coordinate, the row classification information corresponding to each seedling coordinate, and the center coordinate of the rice seedling field image includes:
when the average value of the first axial coordinate values of all the seedling coordinates is larger than the first axial coordinate value of the central coordinate of the field image of the rice seedlings, confirming the row classification information corresponding to the seedling coordinate with the minimum first axial coordinate value as the row classification information of the seedling row in the current operation;
and when the average value of the first axial coordinate values of the seedling coordinates is smaller than the first axial coordinate value of the central coordinate of the field image of the rice seedlings, determining the row classification information corresponding to the seedling coordinate with the largest first axial coordinate value as the row classification information of the seedling row in the current operation.
In one embodiment, the process of performing a linear regression of the seedling coordinates of the currently operated seedling row to obtain the current seedling row line comprises:
obtaining each seedling coordinate on the seedling row in the current operation according to each seedling coordinate, the row classification information corresponding to each seedling coordinate and the row classification information of the seedling row in the current operation;
and fitting each seedling coordinate on the seedling row in the current operation through a linear regression algorithm to obtain the current seedling row line.
On the other hand, the embodiment of the invention also provides a device for extracting the row lines of the rice seedlings, which comprises:
the image acquisition module is used for acquiring a rice seedling field image;
the seedling coordinate recognition module is used for recognizing the rice seedling field image through a pre-trained deep convolutional neural network model to obtain each seedling coordinate in the rice seedling field image;
the current operation seedling row identification module is used for clustering a plurality of seedling coordinates and identifying a current operation seedling row;
and the seedling row line acquisition module is used for performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line.
In another aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and the processor implements the following steps when executing the computer program:
acquiring a rice seedling field image;
recognizing the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image;
clustering a plurality of seedling coordinates, and identifying the seedling row in current operation;
and performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line.
In one aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the following steps:
acquiring a rice seedling field image;
recognizing the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image;
clustering a plurality of seedling coordinates, and identifying the seedling row in current operation;
and performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line.
According to the method, the device, the computer equipment and the storage medium for extracting the rice seedling row lines, the seedling coordinates are identified through the pre-trained deep convolutional neural network model, the seedling coordinates are clustered, the seedling coordinates are classified according to the seedling rows, the current seedling row lines are obtained through identifying the current operation seedling rows, and the navigation is carried out on the rice transplanter. Based on the method, through the application of the deep convolutional neural network and the clustering algorithm, the interference of environmental factors in the rice field is better eliminated, the current rice seedling row line can be more accurately extracted, and the automatic driving of the rice transplanter is guided.
Drawings
FIG. 1 is a schematic flow chart of a method for extracting rice seedling row lines in one embodiment;
FIG. 2 is a schematic flow chart of a method for row line extraction of rice seedlings according to another embodiment;
FIG. 3 is a schematic flow chart of the process for clustering multiple seedling coordinates and identifying currently operating seedling rows in one embodiment;
FIG. 4 is a schematic flow chart of the process for clustering multiple seedling coordinates and identifying the currently operating seedling row in another embodiment;
FIG. 5 is a schematic flow chart illustrating the process of performing linear regression on seedling coordinates of currently operated seedling rows to obtain current seedling row lines in one embodiment;
FIG. 6 is a block diagram of a row line extractor for rice seedlings in one embodiment;
FIG. 7 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
As shown in fig. 1, an embodiment of the present invention provides a method for extracting rice seedling row lines, including:
step S30, acquiring a rice seedling field image;
step S40, recognizing the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image;
step S50, clustering a plurality of seedling coordinates, and identifying the seedling row in current operation;
and step S60, performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line.
Wherein the seedling coordinates are pixel point position coordinates of the seedling plants on the rice seedling field image. The current operation seedling row is the seedling row which is currently transplanted by the transplanter.
Specifically, a pre-trained deep convolutional neural network model is used for identifying each seedling coordinate in the rice seedling field image; clustering the seedling rows of the plurality of recognized seedling coordinates, and recognizing the seedling row in the plurality of clustered seedling rows in the current operation; and acquiring coordinates of seedlings in the seedling row of the current operation, and performing linear regression to obtain a current seedling row line.
According to the method for extracting the rice seedling row line, the seedling coordinates are identified through a pre-trained deep convolutional neural network model, the seedling coordinates are clustered, the seedling coordinates are classified according to the seedling lines, the current seedling row line is obtained through identifying the current operation seedling line, and navigation is performed on the rice transplanter. Based on the method, through the application of the deep convolutional neural network and the clustering algorithm, the interference of environmental factors in the rice field is better eliminated, the current rice seedling row line can be more accurately identified, and the automatic driving of the rice transplanter is guided.
In one embodiment, as shown in fig. 2, before acquiring the image of the rice seedling field, the method further includes:
step S10, acquiring sample data of the rice seedling field image, wherein the sample data of the rice seedling field image comprises rice seedling field images of various natural conditions, illumination angles and rice varieties;
and step S20, training the deep convolution neural network model according to the sample data of the rice seedling field image.
Specifically, the rice seedling field image sample data includes image samples under various conditions, including various natural conditions (e.g., sunny days, cloudy days, rainy days, etc.), various illumination angles (e.g., lighting, backlighting, photometry, etc.), and various rice varieties. The method comprises the steps of inputting rice seedling field image samples under various different conditions into a deep convolution neural network model, training the deep convolution neural network model, and enabling parameters of the deep convolution neural network model to be updated in an iterative mode, so that the recognition capability of the deep convolution neural network model on the rice seedling field images under different conditions is enhanced, and the recognition of seedling coordinates is more accurate.
In one embodiment, the deep convolutional neural network model is a fast RCNN neural network model. Compared with other deep convolution neural network models, the FasterRCNN neural network model has better adaptability to different images and can accurately complete the identification of seedling coordinates.
In one embodiment, the training process of the Faster RCNN neural network model based on the sample data of the rice seedling field image includes:
pre-training a Faster RCNN neural network model by applying a ZF neural network model;
and training an RPN neural network model and a FastCNN neural network model in a Faster RCNN neural network model by adopting an alternative training mode. In the alternate training process, the parameters of the RPN neural network model and the Fast RCNN neural network model are optimized and updated by using a random gradient descent algorithm, and the learning rate of the random gradient descent algorithm is 0.001.
In one embodiment, the process of recognizing the rice seedling field image through the pre-trained fast RCNN neural network model to obtain each seedling coordinate in the rice seedling field image includes:
performing convolution feature extraction on the rice seedling field image through a pre-trained Fast RCNN neural network model to obtain a convolution feature map of the rice seedling field image;
identifying the convolution characteristic graph of the rice seedling field image through a pre-trained RPN neural network model to obtain a plurality of seedling candidate frames;
identifying each seedling area frame in each seedling candidate frame by using a pre-trained Fast RCNN neural network model;
and calculating to obtain the coordinates of each seedling according to each rice plant area frame.
Wherein the seedling candidate frame and the seedling area frame are rectangular areas on a convolution characteristic diagram of the rice seedling field image.
Specifically, the Fast RCNN neural network model includes a Fast RCNN neural network model and an RPN neural network model. Matching the Fast RCNN neural network model with the RPN neural network model to respectively complete some steps in the seedling coordinate identification process.
And performing convolution feature extraction on the rice seedling field image through a pre-trained Fast RCNN neural network model to obtain a convolution feature map of the rice seedling field image. And identifying a plurality of seedling candidate frames on the convolution characteristic graph of the rice seedling field image by using a pre-trained RPN neural network model to obtain a plurality of seedling candidate frame data. The seedling candidate frame data comprises: coordinates of a starting point of the seedling candidate frame, a first side length of the candidate frame, a second side length of the candidate frame and image data in the candidate frame. The starting point of the seedling candidate frame is a pixel point with the smallest horizontal and vertical coordinates on the seedling candidate frame. And recognizing the seedling area frame in each seedling candidate frame image by using a pre-trained Fast RCNN neural network model to obtain a plurality of seedling area frame data. The seedling area frame data includes: the coordinates of the starting point of the seedling area frame, the first side length of the area frame, the second side length of the area frame and the image data in the area frame. The starting point of the seedling area frame is a pixel point with the smallest horizontal and vertical coordinates on the seedling area frame. And calculating the coordinates of the central point of each seedling area frame according to the coordinates of the initial point of each seedling area frame data, the first side length of the area frame and the second side length of the area frame, and taking the coordinates of the central point of each seedling area frame as the coordinates of each seedling.
In one embodiment, as shown in fig. 3, clustering the plurality of seedling coordinates and identifying the currently operating seedling row includes:
step S51, classifying the seedling coordinates through a clustering algorithm to obtain row classification information corresponding to each seedling coordinate;
and step S52, identifying the seedling row currently operated according to each seedling coordinate, the row classification information corresponding to each seedling coordinate and the central coordinate of the rice seedling field image.
Wherein, the point where the central lines of all axial directions of the rice seedling field image intersect is the central point of the rice seedling field image, and the coordinate of the central point is the central coordinate of the rice seedling field image.
Specifically, the seedling coordinates are classified according to the rows by using a clustering algorithm according to all the seedling coordinates, and row classification information of each seedling coordinate is obtained. And calculating the line classification information of the seedling line in the current operation according to the seedling coordinates, the line classification information corresponding to the seedling coordinates and the central coordinates of the field image of the rice seedlings, and determining the seedling line in the current operation. In one embodiment, the clustering algorithm may be a K-means clustering algorithm or a mean-shift clustering algorithm, or the like. Preferably, the clustering algorithm is a hierarchical clustering algorithm.
In one embodiment, as shown in fig. 4, the process of identifying the currently operating seedling row based on each seedling coordinate, the row classification information corresponding to each seedling coordinate, and the center coordinate of the rice seedling field image includes:
step S521, when the average value of the first axial coordinate values of the seedling coordinates is larger than the first axial coordinate value of the central coordinate of the field image of the rice seedlings, confirming the row classification information corresponding to the seedling coordinate with the minimum first axial coordinate value as the row classification information of the seedling row in the current operation;
step S522, when the average value of the first axial coordinate values of the seedling coordinates is smaller than the first axial coordinate value of the central coordinate of the field image of the rice seedling, the row classification information corresponding to the seedling coordinate with the largest first axial coordinate value is determined and is the row classification information of the seedling row in the current operation.
The first axial coordinate value may be a lateral coordinate value (X-axis value) of the coordinate point, or may be a vertical coordinate value (Y-axis value) of the coordinate point.
Specifically, taking the first axial coordinate value as the lateral coordinate value of the coordinate point as an example, the following explanation is given:
calculating the average value of the transverse coordinate values of each seedling coordinate, wherein the calculation process is as follows:
wherein n is the number of seedling coordinates.
Comparing the average value of the transverse coordinate values of each seedling coordinate
And a transverse coordinate value of a central coordinate of the rice seedling field image;
when each seedling coordinate is horizontalAverage value of coordinate values
If the horizontal coordinate value is larger than the central coordinate of the field image of the rice seedling, searching the seedling coordinate with the minimum horizontal coordinate value in the horizontal coordinate values of the seedling coordinates, and determining the classification information corresponding to the seedling coordinate as the row classification information of the seedling row in the current operation;
when the average value of the transverse coordinate values of each seedling coordinate
And if the horizontal coordinate value is smaller than the central coordinate of the field image of the rice seedling, searching the seedling coordinate with the maximum horizontal coordinate value in the horizontal coordinate values of the seedling coordinates, and determining the classification information corresponding to the seedling coordinate as the row classification information of the seedling row in the current operation.
In one embodiment, as shown in fig. 5, the process of performing a linear regression of the seedling coordinates on the currently operated seedling row to obtain the current seedling row line comprises:
step S61, obtaining each seedling coordinate on the seedling row in the current operation according to each seedling coordinate, the row classification information corresponding to each seedling coordinate and the row classification information of the seedling row in the current operation;
and step S62, fitting each seedling coordinate on the current operation seedling row through a linear regression algorithm to obtain the current seedling row.
Specifically, the seedling coordinates corresponding to the row classification information of the currently operated seedling row, that is, the seedling coordinates on the currently operated seedling row, are found according to the seedling coordinates and the row classification information corresponding to the seedling coordinates. And performing linear regression fitting on each seedling coordinate on the current operation seedling row to obtain a current seedling row line, and guiding the navigation of the rice transplanter.
On the other hand, the embodiment of the invention also provides a device for extracting the row line of rice seedlings, which is shown in fig. 6 and comprises:
the image acquisition module 710 is used for acquiring a rice seedling field image;
the seedling coordinate recognition module 720 is used for recognizing the rice seedling field image through a pre-trained deep convolutional neural network model to obtain each seedling coordinate in the rice seedling field image;
the current operation seedling row identification module 730 is used for clustering a plurality of seedling coordinates and identifying a current operation seedling row;
and the seedling row line obtaining module 740 is used for performing linear regression on the seedling coordinates of the seedling row in the current operation to obtain the current seedling row line.
For the specific limitations of the device for extracting the rice seedling row line, reference can be made to the above limitations of the method for extracting the rice seedling row line, and the details are not repeated here. All modules in the rice seedling row line extraction device can be completely or partially realized through software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a personal computer, the internal structure of which may be as shown in fig. 7. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer equipment is used for storing sample data of the rice seedling field image. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to realize a rice seedling row line extraction method.
Those skilled in the art will appreciate that the architecture shown in fig. 7 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring a rice seedling field image;
identifying the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image;
clustering a plurality of seedling coordinates, and identifying the seedling row in current operation;
and performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line.
In one embodiment, the invention provides a rice seedling row line extraction system which comprises an image acquisition device and a data processing device.
The image acquisition equipment is electrically connected with the data processing equipment and is used for acquiring the field images of the rice seedlings. The image acquisition equipment is arranged at the rear end of the rice transplanter and used for acquiring the field image of the rice seedlings on the back of the rice transplanter and sending the field image to the data processing equipment when the rice transplanter moves forwards for transplanting. The data processing equipment receives the field image of the rice seedlings and executes the rice seedling row line extraction method.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring a rice seedling field image;
identifying the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image;
clustering a plurality of seedling coordinates, and identifying the seedling row in current operation;
and performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain the current seedling row line.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method for extracting rice seedling row lines is characterized by comprising the following steps:
acquiring a rice seedling field image;
identifying the rice seedling field image through a pre-trained deep convolution neural network model to obtain each seedling coordinate in the rice seedling field image;
clustering a plurality of seedling coordinates, and identifying a current operation seedling row;
performing linear regression on the seedling coordinates on the current operation seedling row to obtain a current seedling row line;
wherein, clustering a plurality of the seedling coordinates and identifying the seedling row currently operated comprises:
classifying the seedling coordinates through a clustering algorithm to obtain row classification information corresponding to each seedling coordinate;
when the average value of the first axial coordinate values of the seedling coordinates is larger than the first axial coordinate value of the central coordinate of the rice seedling field image, confirming the row classification information corresponding to the seedling coordinate with the minimum first axial coordinate value as the row classification information of the seedling row in the current operation;
and when the average value of the first axial coordinate values of the seedling coordinates is smaller than the first axial coordinate value of the central coordinate of the rice seedling field image, confirming the row classification information corresponding to the seedling coordinate with the maximum first axial coordinate value as the row classification information of the seedling row in the current operation.
2. The method of claim 1, further comprising, prior to obtaining the field image of the rice seedling:
acquiring sample data of a field image of the rice seedling, wherein the sample data of the field image of the rice seedling comprises field images of the rice seedling of various natural conditions, illumination angles and varieties;
and training a deep convolution neural network model according to the sample data of the rice seedling field image.
3. The rice seedling row line extraction method as claimed in claim 2, wherein the deep convolutional neural network model is a fast RCNN neural network model.
4. The rice seedling row line extraction method as claimed in claim 3, wherein the Fast RCNN neural network model comprises a Fast RCNN neural network model and an RPN neural network model.
5. The method of claim 4, wherein the identifying the field image of the rice seedling through the pre-trained deep convolutional neural network model to obtain the coordinates of each seedling in the field image of the rice seedling comprises:
performing convolution feature extraction on the rice seedling field image through the Fast RCNN neural network model trained in advance to obtain a convolution feature map of the rice seedling field image;
identifying the convolution characteristic graph of the rice seedling field image through a preset trained RPN neural network neural model to obtain a plurality of seedling candidate frames;
identifying each seedling area frame in the plurality of seedling candidate frames by applying the Fast RCNN neural network model;
and obtaining the coordinates of each seedling according to each seedling area frame.
6. The method of claim 1 wherein the clustering algorithm is a hierarchical clustering algorithm.
7. The method of claim 1, wherein the linear regression of the seedling coordinates of the currently operated seedling row to obtain the current seedling row comprises:
obtaining each seedling coordinate on the seedling row in the current operation according to each seedling coordinate, the row classification information corresponding to each seedling coordinate and the row classification information of the seedling row in the current operation;
and fitting each seedling coordinate on the current operation seedling row through a linear regression algorithm to obtain the current seedling row.
8. A rice seedling row line extraction element which characterized in that includes:
the image acquisition module is used for acquiring a rice seedling field image;
the seedling coordinate recognition module is used for recognizing the rice seedling field image through a pre-trained deep convolutional neural network model to obtain each seedling coordinate in the rice seedling field image;
the current operation seedling row identification module is used for clustering a plurality of seedling coordinates and identifying a current operation seedling row;
the seedling row line acquisition module is used for performing linear regression on the seedling coordinates on the seedling row in the current operation to obtain a current seedling row line;
wherein, current operation seedling is gone to identification module includes:
the line classification information determining unit is used for classifying the seedling coordinates through a clustering algorithm to obtain line classification information corresponding to each seedling coordinate;
a first judging unit, configured to, when an average value of first axial coordinate values of the seedling coordinates is greater than a first axial coordinate value of a center coordinate of the rice seedling field image, determine row classification information corresponding to the seedling coordinate with the smallest first axial coordinate value as row classification information of the currently operated seedling row;
and the second judgment unit is used for confirming the row classification information corresponding to the seedling coordinate with the maximum first axial coordinate value as the row classification information of the seedling row currently operated when the average value of the first axial coordinate values of all the seedling coordinates is smaller than the first axial coordinate value of the central coordinate of the rice seedling field image.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor when executing the computer program implements the steps of the row line extraction method for rice seedlings according to any one of claims 1 to 7.
10. A computer-readable storage medium having stored thereon a computer program for implementing the steps of the method for row line extraction of rice seedlings according to any one of claims 1 to 7 when the computer program is executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811043572.2A CN109344843B (en) | 2018-09-07 | 2018-09-07 | Method and device for extracting rice seedling row line, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811043572.2A CN109344843B (en) | 2018-09-07 | 2018-09-07 | Method and device for extracting rice seedling row line, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109344843A CN109344843A (en) | 2019-02-15 |
| CN109344843B true CN109344843B (en) | 2020-09-25 |
Family
ID=65304556
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811043572.2A Active CN109344843B (en) | 2018-09-07 | 2018-09-07 | Method and device for extracting rice seedling row line, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109344843B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948543B (en) * | 2019-03-20 | 2023-02-03 | 广西科技大学 | Method for extracting leading line of field close-range image |
| CN113065503B (en) * | 2021-04-15 | 2023-12-22 | 黑龙江惠达科技股份有限公司 | Seedling line identification method and device based on monocular camera and agricultural machinery |
| CN113065501B (en) * | 2021-04-15 | 2024-03-22 | 黑龙江惠达科技股份有限公司 | Seedling line identification method and device and agricultural machinery |
| CN113536958B (en) * | 2021-06-23 | 2023-08-25 | 华南农业大学 | Navigation path extraction method, device, agricultural robot and storage medium |
| CN113435309B (en) * | 2021-06-24 | 2022-04-26 | 农业农村部南京农业机械化研究所 | Rice seedling row identification method based on row vector grid classification |
| CN115451965B (en) * | 2022-08-19 | 2024-04-26 | 安徽农业大学 | Relative heading information detection method for transplanting system of transplanting machine based on binocular vision |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103186773A (en) * | 2013-02-06 | 2013-07-03 | 内蒙古大学 | Early-stage ribbing ridge line recognition algorithm based on one-dimensional Hough transform and expert system |
| CN103646249A (en) * | 2013-12-12 | 2014-03-19 | 江苏大学 | Greenhouse intelligent mobile robot vision navigation path identification method |
| CN107067430A (en) * | 2017-04-13 | 2017-08-18 | 河南理工大学 | A kind of wheatland crop row detection method of distinguished point based cluster |
| CN107169421A (en) * | 2017-04-20 | 2017-09-15 | 华南理工大学 | A kind of car steering scene objects detection method based on depth convolutional neural networks |
| CN107463903A (en) * | 2017-08-08 | 2017-12-12 | 北京小米移动软件有限公司 | Face key independent positioning method and device |
| EP3367648A1 (en) * | 2017-02-24 | 2018-08-29 | Lg Electronics Inc. | Automatic photographing dependent on positioning on preview screen |
-
2018
- 2018-09-07 CN CN201811043572.2A patent/CN109344843B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103186773A (en) * | 2013-02-06 | 2013-07-03 | 内蒙古大学 | Early-stage ribbing ridge line recognition algorithm based on one-dimensional Hough transform and expert system |
| CN103646249A (en) * | 2013-12-12 | 2014-03-19 | 江苏大学 | Greenhouse intelligent mobile robot vision navigation path identification method |
| EP3367648A1 (en) * | 2017-02-24 | 2018-08-29 | Lg Electronics Inc. | Automatic photographing dependent on positioning on preview screen |
| CN107067430A (en) * | 2017-04-13 | 2017-08-18 | 河南理工大学 | A kind of wheatland crop row detection method of distinguished point based cluster |
| CN107169421A (en) * | 2017-04-20 | 2017-09-15 | 华南理工大学 | A kind of car steering scene objects detection method based on depth convolutional neural networks |
| CN107463903A (en) * | 2017-08-08 | 2017-12-12 | 北京小米移动软件有限公司 | Face key independent positioning method and device |
Non-Patent Citations (4)
| Title |
|---|
| Line-based paddy boundary extraction using the RapidEye satellite image;Junho Yeom et al;《2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS)》;20151112;第3397-3400页 * |
| Vision based guidance line extraction for autonomous weed control robot in paddy field;Keun Ha Choi et al;《2015 IEEE International Conference on Robotics and Biomimetics (ROBIO)》;20160225;第3部分A部分第1段、C部分、图6(d)、D部分第1段、第4部分第1段 * |
| 基于机器视觉的小型农业AGV玉米田间路径识别研究;刘永博;《中国优秀硕士学位论文全文数据库 信息科技辑》;20170615(第6期);第I138-1155页 * |
| 基于机器视觉的茶陇识别与采茶机导航方法;汤一平等;《农业机械学报》;20160131;第47卷(第1期);第45-50页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109344843A (en) | 2019-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109344843B (en) | Method and device for extracting rice seedling row line, computer equipment and storage medium | |
| CN109863874B (en) | Fruit and vegetable picking method, picking device and storage medium based on machine vision | |
| CN109241903B (en) | Sample data cleaning method, device, computer equipment and storage medium | |
| CN111950656B (en) | Image recognition model generation method and device, computer equipment and storage medium | |
| CN111553240B (en) | Corn disease condition grading method and system and computer equipment | |
| EP3832524A1 (en) | System and method for determining damage on crops | |
| CN111723863B (en) | Fruit tree flower identification and position acquisition method and device, computer equipment and storage medium | |
| CN111753692A (en) | Target object extraction method, product detection method, device, computer and medium | |
| CN111767849A (en) | Crop pest and disease identification method and device and storage medium | |
| CN112949704B (en) | Tobacco leaf maturity state identification method and device based on image analysis | |
| Li et al. | Real-time crop recognition in transplanted fields with prominent weed growth: a visual-attention-based approach | |
| CN109712128A (en) | Feature point detecting method, device, computer equipment and storage medium | |
| CN108460344A (en) | Dynamic area intelligent identifying system in screen and intelligent identification Method | |
| CN115620151B (en) | Method and device for identifying phenological period, electronic equipment and storage medium | |
| CN111899247B (en) | Method, device, equipment and medium for identifying lumen area of choroidal blood vessel | |
| Kumar et al. | Segmentation of rice seedling using deep learning algorithm | |
| CN117392535A (en) | Fruit tree flower bud target detection and white point rate estimation method oriented to complex environment | |
| CN111696122A (en) | Crop phenotype parameter extraction method and device | |
| CN115375674B (en) | Stomach white-light neoplasia image identification method, device and storage medium | |
| CN114937030A (en) | Phenotypic parameter calculation method for intelligent agricultural planting of lettuce | |
| CN113128514A (en) | Cotton pest positioning and classifying identification method and system | |
| CN109308936B (en) | Grain crop production area identification method, grain crop production area identification device and terminal identification equipment | |
| CN114463642A (en) | Cultivated land plot extraction method based on deep learning | |
| CN111723736A (en) | Fruit tree flowering phase monitoring method and device, computer equipment and storage medium | |
| CN113408374A (en) | Yield estimation method, device and equipment based on artificial intelligence and storage medium |
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 |