CN109344843A - Rice seedling line extracting method, device, computer equipment and storage medium - Google Patents
Rice seedling line extracting method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109344843A CN109344843A CN201811043572.2A CN201811043572A CN109344843A CN 109344843 A CN109344843 A CN 109344843A CN 201811043572 A CN201811043572 A CN 201811043572A CN 109344843 A CN109344843 A CN 109344843A
- Authority
- CN
- China
- Prior art keywords
- rice
- rice shoot
- coordinate
- row
- shoot
- 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.)
- Granted
Links
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 present invention relates to a kind of rice seedling line extracting methods, comprising steps of obtaining rice seedling field image；Rice seedlings field image is identified by preparatory trained depth convolutional neural networks model, obtains each rice shoot coordinate in rice seedling field image；Multiple rice shoot coordinates are clustered, and identify current work rice shoot row；Linear regression is carried out to the rice shoot coordinate on current work rice shoot row, obtains current rice shoot line.By the application of preparatory trained depth convolutional neural networks and clustering algorithm, the environmental factor interference of rice field is preferably excluded, current rice shoot line can be more accurately identified, instruct rice transplanting machine automatic drive.
Description
Technical field
The present invention relates to field of image processings, more particularly to a kind of rice seedling line extracting method, device, computer
Equipment and storage medium.
Background technique
Agriculture Field airmanship provides navigation information for the operation of agricultural modernization manufacturing machine, is applied to each
In kind agricultural machinery.In the agricultural equipment for sowing, the importance of Agriculture Field navigation is particularly evident.
Currently, in terms of the research of Agriculture Field navigation both at home and abroad is concentrated mainly on machine vision navigation.Meng Qing wide et al. structure
The Cg component unrelated with illumination is built, selects 2Cg-Cr-Cb characterization factor to carry out gray processing processing to image, according to crop in image
Capable feature establishes the restricted model of corn crop row linear equation, carries out optimizing to crop row straight line using particle swarm algorithm and asks
Solution, and then obtain leading line.Picture contrast caused by a variety of illumination conditions is unobvious or supersaturation to adapt to, Romeo
Etc. devising a kind of crop based on image histogram analysis-background image segmenting system, which passes through histogram and differentiates figure
The contrast and saturation degree of picture select suitable wheat seedling image partition method.
The accuracy that accurately identifies with navigation line drawing of the seedling with image information, is before realizing the accurate operation of implanted device
It mentions.Rice seedling line is identified in the traditional machine vision method of application, during the realization for rice transplanter navigation, invention
At least there are the following problems for people's discovery: rice transplanting stage rice field environment is complicated, and water layer is relatively thin, and traditional image-recognizing method is very
Difficult accurately identification rice seedling line, instructs rice transplanter autonomous driving.
Summary of the invention
Based on this, it is necessary to identify not accurate enough problem for rice seedling line, provide a kind of rice seedling line
Extracting method, device, computer equipment and storage medium.
On the one hand, the embodiment of the present invention provides a kind of rice seedling line extracting method, comprising:
Obtain rice seedling field image；
Rice seedling field image is identified by preparatory trained depth convolutional neural networks model, obtains water
Each rice shoot coordinate in rice sprouts field image；
Multiple rice shoot coordinates are clustered, and identify current work rice shoot row；
Linear regression is carried out to the rice shoot coordinate on current work rice shoot row, obtains current rice shoot line.
In one of the embodiments, before acquisition rice seedling field image, further includes:
The sample data of rice seedling field image is obtained, the sample data of rice seedling field image includes a variety of natures
The rice seedling field image of condition, lighting angle and kind；
Depth convolutional neural networks model is trained according to the sample data of rice seedling field image.
Depth convolutional neural networks model is Faster RCNN neural network model in one of the embodiments,.
Multiple rice shoot coordinates are clustered in one of the embodiments, identify the process packet of current work rice shoot row
It includes:
Classified by clustering algorithm to multiple rice shoot coordinates, obtains the corresponding row classification information of each rice shoot coordinate；
According to each rice shoot coordinate, the center of each rice shoot coordinate corresponding row classification information and rice seedling field image
Coordinate identifies current work rice shoot row.
Clustering algorithm is hierarchical clustering algorithm in one of the embodiments,.
In one of the embodiments, according to each rice shoot coordinate, the corresponding row classification information of each rice shoot coordinate and water
The centre coordinate of rice sprouts field image identifies that the process of current work rice shoot row includes:
When the average value of the first axis coordinate value of each rice shoot coordinate is greater than the centre coordinate of rice seedling field image
When first axis coordinate value, the corresponding row classification information of the confirmation the smallest rice shoot coordinate of first axis coordinate value is current work
The row classification information of rice shoot row；
When the average value of the first axis coordinate value of each rice shoot coordinate is less than the centre coordinate of rice seedling field image
When first axis coordinate value, the corresponding row classification information of the confirmation maximum rice shoot coordinate of first axis coordinate value is current work
The row classification information of rice shoot row.
Linear regression is carried out to the rice shoot coordinate on current work rice shoot row in one of the embodiments, is obtained current
The process of rice shoot line includes:
According to the row classification of each rice shoot coordinate, the corresponding row classification information of each rice shoot coordinate and current work rice shoot row
Information obtains each rice shoot coordinate on current work rice shoot row；
By linear regression algorithm, each rice shoot coordinate on current work rice shoot row is fitted, current rice shoot is obtained
Line.
On the other hand, the embodiment of the present invention also provides a kind of rice seedling line extraction element, comprising:
Image collection module, for obtaining rice seedling field image；
Rice shoot coordinate identification module, for passing through preparatory trained depth convolutional neural networks model to rice seedling field
Between image identified, obtain each rice shoot coordinate in rice seedling field image；
Current work rice shoot row identification module for clustering to multiple rice shoot coordinates, and identifies current work rice shoot
Row；
Rice shoot line obtains module, for carrying out linear regression to the rice shoot coordinate on current work rice shoot row, is worked as
Preceding rice shoot line.
In another aspect, the embodiment of the present invention provides a kind of computer equipment, including memory and processor, the memory
It is stored with computer program, the processor performs the steps of when executing the computer program
Obtain rice seedling field image；
Rice seedling field image is identified by preparatory trained depth convolutional neural networks model, obtains water
Each rice shoot coordinate in rice sprouts field image；
Multiple rice shoot coordinates are clustered, and identify current work rice shoot row；
Linear regression is carried out to the rice shoot coordinate on current work rice shoot row, obtains current rice shoot line.
On the one hand, the embodiment of the present invention provides a kind of computer readable storage medium, is stored thereon with computer program, institute
It states when computer program is executed by processor and performs the steps of
Obtain rice seedling field image；
Rice seedling field image is identified by preparatory trained depth convolutional neural networks model, obtains water
Each rice shoot coordinate in rice sprouts field image；
Multiple rice shoot coordinates are clustered, and identify current work rice shoot row；
Linear regression is carried out to the rice shoot coordinate on current work rice shoot row, obtains current rice shoot line.
Above-mentioned rice seedling line extracting method, device, computer equipment and storage medium pass through preparatory trained depth
It spends convolutional neural networks model and identifies rice shoot coordinate, rice shoot coordinate is clustered, rice shoot coordinate is pressed into the classification of rice shoot row, passes through identification
Current work rice shoot row obtains current rice shoot line, navigates for rice transplanter.Based on this, pass through depth convolutional Neural
The application of network and clustering algorithm preferably excludes the environmental factor interference of rice field, can more accurately extract current seedling
Seedling line instructs rice transplanting machine automatic drive.
Detailed description of the invention
Fig. 1 is the flow diagram of rice seedling line extracting method in an embodiment；
Fig. 2 is the flow diagram of rice seedling line extracting method in another embodiment；
Fig. 3 is to cluster in an embodiment to multiple rice shoot coordinates, and the process of identification current work rice shoot row step is shown
It is intended to；
Fig. 4 is to cluster in another embodiment to multiple rice shoot coordinates, identifies the process of current work rice shoot row step
Schematic diagram；
Fig. 5 is to carry out linear regression to the rice shoot coordinate on current work rice shoot row in an embodiment, obtains current rice shoot
The flow diagram of line step；
Fig. 6 is the structural block diagram of rice seedling line extraction element in an embodiment；
Fig. 7 is the internal structure chart of computer equipment in an embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
As shown in Figure 1, the embodiment of the present invention provides a kind of rice seedling line extracting method, comprising:
Step S30 obtains rice seedling field image；
Step S40 identifies rice seedlings field image by preparatory trained depth convolutional neural networks model,
Obtain each rice shoot coordinate in rice seedling field image；
Step S50 clusters multiple rice shoot coordinates, and identifies current work rice shoot row；
Step S60 carries out linear regression to the rice shoot coordinate on current work rice shoot row, obtains current rice shoot line.
Wherein, rice shoot coordinate is pixel position coordinates of the rice shoot seedling strain on rice seedling field image.Current work
Rice shoot row is rice transplanter currently just in the rice shoot row of rice transplanting.
Specifically, being identified in rice seedling field image with preparatory trained depth convolutional neural networks model
Each rice shoot coordinate；The cluster that the multiple rice shoot coordinates recognized are carried out to rice shoot row is known in multiple rice shoot rows that cluster obtains
It is clipped to current work rice shoot row；The coordinate of the rice shoot in current work rice shoot row is obtained, and carries out linear regression, obtains current seedling
Seedling line.
Rice seedling line extracting method provided in this embodiment passes through preparatory trained depth convolutional neural networks mould
Type identifies rice shoot coordinate, clusters to rice shoot coordinate, and rice shoot coordinate is pressed the classification of rice shoot row, by identifying current work rice shoot row,
Current rice shoot line is obtained, is navigated for rice transplanter.Based on this, pass through depth convolutional neural networks and clustering algorithm
Using the environmental factor for preferably excluding rice field is interfered, and can more accurately be identified current rice shoot line, be instructed rice transplanter
Automatic Pilot.
In one embodiment, as shown in Fig. 2, before obtaining rice seedling field image, further includes:
Step S10, obtains the sample data of rice seedling field image, and the sample data of rice seedling field image includes
The rice seedling field image of a variety of natural conditions, lighting angle and rice varieties；
Step S20 is trained depth convolutional neural networks model according to the sample data of rice seedling field image.
Specifically, rice seedling field image sample data include it is a variety of under the conditions of image pattern, including a variety of natures
Condition (such as fine day, cloudy day, rainy day etc.), a variety of lighting angles (such as to light, backlight, survey light etc.) and a variety of rice product
Kind.Rice seedling field image sample under a variety of different conditions is input in depth convolutional neural networks model, to depth
Convolutional neural networks model is trained, and updates the parameter iteration of depth convolutional neural networks model, to enhance depth
Convolutional neural networks model keeps the identification of rice shoot coordinate more quasi- the recognition capability of rice seedling field image under different condition
Really.
In one embodiment, depth convolutional neural networks model is Faster RCNN neural network model.Faster
RCNN neural network model, it is more preferable to the adaptability of different images compared with other depth convolutional neural networks models, it can be accurate
Completion rice shoot coordinate identification.
In one embodiment, according to the sample data of rice seedling field image to Faster RCNN neural network mould
The process that type is trained, comprising:
Pre-training is carried out to Faster RCNN neural network model with ZF neural network model；
Using alternately training method, the RPN neural network model and Fast in Faster RCNN neural network model are trained
RCNN neural network model.In alternately training process, with stochastic gradient descent algorithm, to RPN neural network model and
The parameter of Fast RCNN neural network model optimizes update, and the learning rate of stochastic gradient descent algorithm is 0.001.
In one embodiment, by preparatory trained Faster RCNN neural network model to rice seedling field
Image is identified that the process for obtaining each rice shoot coordinate in rice seedling field image includes:
Convolution feature is carried out to rice seedling field image by preparatory trained Fast RCNN neural network model to mention
It takes, obtains the convolution characteristic pattern of rice seedling field image；
The convolution characteristic pattern of rice seedling field image is known by preparatory trained RPN neural network model
Not, multiple rice shoot candidate frames are obtained；
Each rice shoot area is identified in each rice shoot candidate frame with preparatory trained Fast RCNN neural network model
Domain frame；
According to each rice strain regional frame, each rice shoot coordinate is calculated.
Wherein, rice shoot candidate frame and rice shoot regional frame are the rectangle on the convolution characteristic pattern of rice seedling field image
Domain.
Specifically, Faster RCNN neural network model includes Fast RCNN neural network model and RPN neural network
Model.Fast RCNN neural network model and the cooperation of RPN neural network model, are respectively completed in rice shoot coordinate identification process
Some steps.
Convolution feature is carried out to rice seedling field image by preparatory trained Fast RCNN neural network model to mention
It takes, obtains the convolution characteristic pattern of rice seedling field image.With preparatory on the convolution characteristic pattern of rice seedling field image
Trained RPN neural network model identifies multiple rice shoot candidate frames, obtains multiple rice shoot candidate frame datas.Rice shoot candidate frame number
According to include: rice shoot candidate frame starting point coordinate, the first side length of candidate frame, the second side length of candidate frame and candidate frame in picture number
According to.The starting point of rice shoot candidate frame is all the smallest pixel of transverse and longitudinal coordinate on rice shoot candidate frame.With trained in advance
Fast RCNN neural network model identifies rice shoot regional frame in each rice shoot candidate frame image, obtains multiple rice shoot regional frames
Data.Rice shoot regional frame data include: the starting point coordinate of rice shoot regional frame, the first side length of regional frame, the second side length of regional frame
And image data in regional frame.The starting point of rice shoot regional frame is all the smallest pixel of transverse and longitudinal coordinate on rice shoot regional frame.
According to the second side length of the starting point coordinate of each rice shoot region frame data, the first side length of regional frame and regional frame, each seedling is calculated
The center point coordinate of seedling regional frame, using the center point coordinate of each rice shoot regional frame as each rice shoot coordinate.
In one embodiment, as shown in figure 3, being clustered to multiple rice shoot coordinates, identification current work rice shoot row
Process includes:
Step S51 classifies to rice shoot coordinate by clustering algorithm, obtains the corresponding row classification letter of each rice shoot coordinate
Breath；
Step S52, according to each rice shoot coordinate, the corresponding row classification information of each rice shoot coordinate and rice seedling field image
Centre coordinate, identify current work rice shoot row.
Wherein, the point of each axial middle line intersection of rice seedling field image is the center of rice seedling field image
Point, the coordinate of the central point are the centre coordinate of rice seedling field image.
Specifically, rice shoot coordinate is obtained each rice shoot and is sat by row classification with clustering algorithm according to all rice shoot coordinates
Row classification information where marking.According to each rice shoot coordinate, the corresponding row classification information of each rice shoot coordinate and rice seedling field
The row classification information of current work rice shoot row is calculated in the centre coordinate of image, determines current work rice shoot row.In a reality
It applies in example, clustering algorithm can be K mean cluster algorithm or the elegant clustering algorithm of mean value etc..Preferably, clustering algorithm is level
Clustering algorithm.
In one embodiment, as shown in figure 4, according to each rice shoot coordinate, the corresponding row classification letter of each rice shoot coordinate
The centre coordinate of breath and rice seedling field image identifies that the process of current work rice shoot row includes:
Step S521, when the average value of the first axis coordinate value of each rice shoot coordinate is greater than in rice seedling field image
When the first axis coordinate value of heart coordinate, the corresponding row classification information of the confirmation the smallest rice shoot coordinate of first axis coordinate value is
The row classification information of current work rice shoot row；
Step S522, when the average value of the first axis coordinate value of each rice shoot coordinate is less than in rice seedling field image
When the first axis coordinate value of heart coordinate, the corresponding row classification information of the confirmation maximum rice shoot coordinate of first axis coordinate value is
The row classification information of current work rice shoot row.
Wherein, first axis coordinate value can be the lateral coordinates value (value of X-axis) of coordinate points, be also possible to coordinate points
Longitudinal coordinate value (value of Y-axis).
Specifically, having following explanation so that first axis coordinate value is the lateral coordinates value of coordinate points as an example:
The average value of the lateral coordinates value of each rice shoot coordinate is calculated, calculating process such as following formula:
Wherein, n is the number of rice shoot coordinate.
The average value of the lateral coordinates value of more each rice shoot coordinateWith the cross of the centre coordinate of rice seedling field image
To coordinate value；
When the average value of the lateral coordinates value of each rice shoot coordinateGreater than the cross of the centre coordinate of rice seedling field image
To coordinate value, then in the lateral coordinates value of searching each rice shoot coordinate, lateral coordinates are worth the smallest rice shoot coordinate, determine that the rice shoot is sat
Mark the row classification information that corresponding classification information is current work rice shoot row；
When the average value of the lateral coordinates value of each rice shoot coordinateLess than the cross of the centre coordinate of rice seedling field image
To coordinate value, then in the lateral coordinates value of searching each rice shoot coordinate, lateral coordinates are worth maximum rice shoot coordinate, determine that the rice shoot is sat
Mark the row classification information that corresponding classification information is current work rice shoot row.
In one embodiment, it as shown in figure 5, carrying out linear regression to the rice shoot coordinate on current work rice shoot row, obtains
Process to current rice shoot line includes:
Step S61, according to each rice shoot coordinate, the corresponding row classification information of each rice shoot coordinate and current work rice shoot row
Row classification information, obtain each rice shoot coordinate on current work rice shoot row；
Step S62 is fitted each rice shoot coordinate on current work rice shoot row, is obtained by linear regression algorithm
Current rice shoot line.
Specifically, finding corresponding current work according to each rice shoot coordinate and the corresponding row classification information of each rice shoot coordinate
Each rice shoot coordinate on the rice shoot coordinate of the row classification information of rice shoot row, that is, current work rice shoot row.By current work seedling
Each rice shoot coordinate linear regression fit on seedling row, obtains current rice shoot line, instructs the navigation of rice transplanter.
On the other hand, the embodiment of the present invention also provides a kind of rice seedling line extraction element, as shown in Figure 6, comprising:
Image collection module 710, for obtaining rice seedling field image；
Rice shoot coordinate identification module 720, for passing through preparatory trained depth convolutional neural networks model to rice seedling
Seedling field image is identified, each rice shoot coordinate in rice seedling field image is obtained；
Current work rice shoot row identification module 730 for clustering to multiple rice shoot coordinates, and identifies current work seedling
Miao Hang；
Rice shoot line obtains module 740, for carrying out linear regression to the rice shoot coordinate on current work rice shoot row, obtains
Current rice shoot line.
Specific restriction about rice seedling line extraction element may refer to extract above for rice seedling line
The restriction of method, details are not described herein.Modules in above-mentioned rice seedling line extraction element can be fully or partially through
Software, hardware and combinations thereof are realized.Above-mentioned each module can be embedded in the form of hardware or independently of the place in computer equipment
It manages in device, can also be stored in a software form in the memory in computer equipment, in order to which processor calls execution or more
The corresponding operation of modules.
In one embodiment, a kind of computer equipment is provided, which can be PC, inside
Structure chart can be as shown in Figure 7.The computer equipment includes processor, the memory, network interface connected by system bus
And database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory of the computer equipment
Including non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program sum number
According to library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The meter
The database for calculating machine equipment is used to store the sample data of rice seedling field image.The network interface of the computer equipment is used for
It is communicated with external terminal by network connection.To realize a kind of rice seedling line when the computer program is executed by processor
Extracting method.
It will be understood by those skilled in the art that structure shown in Fig. 7, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor perform the steps of when executing computer program
Obtain rice seedling field image；
Rice seedlings field image is identified by preparatory trained depth convolutional neural networks model, obtains rice
Each rice shoot coordinate in rice shoot field image；
Multiple rice shoot coordinates are clustered, and identify current work rice shoot row；
Linear regression is carried out to the rice shoot coordinate on current work rice shoot row, obtains current rice shoot line.
In one embodiment, the embodiment of the present invention provides a kind of rice seedling line extraction system, including Image Acquisition
Equipment and data processing equipment.
Image capture device is electrically connected with data processing equipment, for acquiring rice seedling field image.Image Acquisition is set
The standby rear end for being mounted on rice transplanter, for acquiring the rice transplanter back side when rice transplanter moves forwards rice transplanting
Rice seedling field image, and it is sent to data processing equipment.Data processing equipment receives rice seedling field image, in execution
The rice seedling line extracting method stated.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of when being executed by processor
Obtain rice seedling field image；
Rice seedlings field image is identified by preparatory trained depth convolutional neural networks model, obtains rice
Each rice shoot coordinate in rice shoot field image；
Multiple rice shoot coordinates are clustered, and identify current work rice shoot row；
Linear regression is carried out to the rice shoot coordinate on current work rice shoot row, obtains current rice shoot line.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to protection of the invention
Range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (10)
1. a kind of rice seedling line extracting method characterized by comprising
Obtain rice seedling field image；
The rice seedling field image is identified by preparatory trained depth convolutional neural networks model, obtains institute
State each rice shoot coordinate in rice seedling field image；
Multiple rice shoot coordinates are clustered, and identify current work rice shoot row；
Linear regression is carried out to the rice shoot coordinate on the current work rice shoot row, obtains current rice shoot line.
2. rice seedling line extracting method according to claim 1, which is characterized in that the acquisition rice seedling field
Before image, further includes:
The sample data of rice seedling field image is obtained, the sample data of the rice seedling field image includes a variety of natures
The rice seedling field image of condition, lighting angle and kind；
Depth convolutional neural networks model is trained according to the sample data of the rice seedling field image.
3. rice seedling line extracting method according to claim 2, which is characterized in that the depth convolutional neural networks
Model is Faster RCNN neural network model.
4. according to claim 1 to rice seedling line extracting method described in 3 any one, which is characterized in that described to more
A rice shoot coordinate is clustered, and the process of identification current work rice shoot row includes:
Classified by clustering algorithm to multiple rice shoot coordinates, obtains the corresponding row classification information of each rice shoot coordinate；
According to each rice shoot coordinate, the corresponding row classification information of each rice shoot coordinate and rice seedling field figure
The centre coordinate of picture identifies current work rice shoot row.
5. rice seedling line extracting method according to claim 4, which is characterized in that the clustering algorithm is poly- for level
Class algorithm.
6. rice seedling line extracting method according to claim 5, which is characterized in that described according to each rice shoot
The centre coordinate of coordinate, each rice shoot coordinate corresponding the row classification information and the rice seedling field image, identification are worked as
The process of preceding operation rice shoot row includes:
When the average value of the first axis coordinate value of each rice shoot coordinate is greater than the centre coordinate of the rice seedling field image
When first axis coordinate value, the corresponding row classification information of the smallest rice shoot coordinate of confirmation first axis coordinate value, is described
The row classification information of current work rice shoot row；
When the average value of the first axis coordinate value of each rice shoot coordinate is less than the centre coordinate of the rice seedling field image
When first axis coordinate value, the corresponding row classification information of the maximum rice shoot coordinate of confirmation first axis coordinate value, is described
The row classification information of current work rice shoot row.
7. rice seedling line extracting method according to claim 6, which is characterized in that described to the current work seedling
Rice shoot coordinate on seedling row carries out linear regression, and the process for obtaining current rice shoot line includes:
According to each rice shoot coordinate, the row of each rice shoot coordinate corresponding the row classification information and the current work rice shoot row
Classification information obtains each rice shoot coordinate on the current work rice shoot row；
By linear regression algorithm, each rice shoot coordinate on the current work rice shoot row is fitted, is obtained described current
Rice shoot line.
8. a kind of rice seedling line extraction element characterized by comprising
Image collection module, for obtaining rice seedling field image；
Rice shoot coordinate identification module, for passing through preparatory trained depth convolutional neural networks model to the rice seedling field
Between image identified, obtain each rice shoot coordinate in the rice seedling field image；
Current work rice shoot row identification module for clustering to multiple rice shoot coordinates, and identifies current work rice shoot
Row；
Rice shoot line obtains module, for carrying out linear regression to the rice shoot coordinate on the current work rice shoot row, is worked as
Preceding rice shoot line.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the processor realizes that rice seedling line described in any one of claims 1 to 7 mentions when executing the computer program
The step of taking method.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of rice seedling line extracting method described in any one of claims 1 to 7 is realized when being executed by 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 true CN109344843A (en) | 2019-02-15 |
| CN109344843B 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) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109948543A (en) * | 2019-03-20 | 2019-06-28 | 广西科技大学 | A kind of leading line extraction method of field close shot image |
| CN110843924A (en) * | 2019-11-29 | 2020-02-28 | 安徽农业大学 | Seedling belt opposite-running auxiliary control method and mechanism based on high-ground-clearance chassis |
| CN113065501A (en) * | 2021-04-15 | 2021-07-02 | 黑龙江惠达科技发展有限公司 | Seedling line identification method and device and agricultural machine |
| CN113065503A (en) * | 2021-04-15 | 2021-07-02 | 黑龙江惠达科技发展有限公司 | Seedling line identification method and device based on monocular camera and agricultural machine |
| CN113435309A (en) * | 2021-06-24 | 2021-09-24 | 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 |
| CN113536958A (en) * | 2021-06-23 | 2021-10-22 | 华南农业大学 | Navigation path extraction method and device, agricultural robot and storage medium |
| CN115451965A (en) * | 2022-08-19 | 2022-12-09 | 安徽农业大学 | Binocular vision-based relative heading information detection method for transplanting system of rice transplanter |
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 |
|---|
| JUNHO YEOM ET AL: "Line-based paddy boundary extraction using the RapidEye satellite image", 《2015 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS)》 * |
| KEUN HA CHOI ET AL: "Vision based guidance line extraction for autonomous weed control robot in paddy field", 《2015 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND BIOMIMETICS (ROBIO)》 * |
| 刘永博: "基于机器视觉的小型农业AGV玉米田间路径识别研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 汤一平等: "基于机器视觉的茶陇识别与采茶机导航方法", 《农业机械学报》 * |
Cited By (13)
| 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 |
| CN109948543A (en) * | 2019-03-20 | 2019-06-28 | 广西科技大学 | A kind of leading line extraction method of field close shot image |
| CN110843924A (en) * | 2019-11-29 | 2020-02-28 | 安徽农业大学 | Seedling belt opposite-running auxiliary control method and mechanism based on high-ground-clearance chassis |
| CN113065501A (en) * | 2021-04-15 | 2021-07-02 | 黑龙江惠达科技发展有限公司 | Seedling line identification method and device and agricultural machine |
| CN113065503A (en) * | 2021-04-15 | 2021-07-02 | 黑龙江惠达科技发展有限公司 | Seedling line identification method and device based on monocular camera and agricultural machine |
| 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 |
| 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 |
| CN113536958B (en) * | 2021-06-23 | 2023-08-25 | 华南农业大学 | Navigation path extraction method, device, agricultural robot and storage medium |
| CN113536958A (en) * | 2021-06-23 | 2021-10-22 | 华南农业大学 | Navigation path extraction method and 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 |
| CN113435309A (en) * | 2021-06-24 | 2021-09-24 | 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 |
| CN115451965A (en) * | 2022-08-19 | 2022-12-09 | 安徽农业大学 | Binocular vision-based relative heading information detection method for transplanting system of rice transplanter |
| 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 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109344843B (en) | 2020-09-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109344843A (en) | Rice seedling line extracting method, device, computer equipment and storage medium | |
| CN111709489B (en) | Citrus identification method based on improved YOLOv4 | |
| CN110148120B (en) | Intelligent disease identification method and system based on CNN and transfer learning | |
| CN109863874B (en) | Fruit and vegetable picking method, picking device and storage medium based on machine vision | |
| Tian et al. | Machine vision identification of tomato seedlings for automated weed control | |
| EP4070230A1 (en) | System and method for determining damage on crops | |
| CN109472801A (en) | It is a kind of for multiple dimensioned neuromorphic detection and dividing method | |
| CN117557787B (en) | Lightweight multi-environment tomato detection method based on improved yolov8 | |
| CN116977960A (en) | Rice seedling row detection method based on example segmentation | |
| CN117036926A (en) | Weed identification method integrating deep learning and image processing | |
| CN117593252A (en) | Crop disease identification method, computer equipment and storage medium | |
| CN117392535A (en) | Fruit tree flower bud target detection and white point rate estimation method oriented to complex environment | |
| CN117152609A (en) | Crop appearance characteristic detecting system | |
| CN116739739A (en) | Loan amount evaluation method and device, electronic equipment and storage medium | |
| CN116311228A (en) | Uncertainty sampling-based corn kernel identification method and system and electronic equipment | |
| CN116188855A (en) | Multi-scale plant disease identification method, device, storage medium and apparatus | |
| CN111160079A (en) | Method for rapidly identifying flowering phase of citrus | |
| Patel et al. | Deep Learning-Based Model for Detection of Brinjal Weed in the Era of Precision Agriculture. | |
| CN113643248B (en) | Wheat fertility process monitoring method based on improved convolutional neural network | |
| CN115451965A (en) | Binocular vision-based relative heading information detection method for transplanting system of rice transplanter | |
| CN113435302A (en) | GridR-CNN-based hydroponic lettuce seedling state detection method | |
| CN112329697A (en) | Improved YOLOv 3-based on-tree fruit identification method | |
| CN114078228A (en) | Target identification method and system based on lightweight network and agricultural machine | |
| AU2021277762B2 (en) | Water level measurement method based on deep convolutional network and random field | |
| CN118411609B (en) | Rice disease seed identification method and system based on random forest |
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 |