CN110400350A - A kind of cane stalk recognition method based on computer vision - Google Patents
A kind of cane stalk recognition method based on computer vision Download PDFInfo
- Publication number
- CN110400350A CN110400350A CN201910654639.4A CN201910654639A CN110400350A CN 110400350 A CN110400350 A CN 110400350A CN 201910654639 A CN201910654639 A CN 201910654639A CN 110400350 A CN110400350 A CN 110400350A
- Authority
- CN
- China
- Prior art keywords
- sugarcane
- value
- stipes
- image
- description vectors
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 36
- 240000000111 Saccharum officinarum Species 0.000 claims abstract description 91
- 235000007201 Saccharum officinarum Nutrition 0.000 claims abstract description 91
- 239000013598 vector Substances 0.000 claims abstract description 44
- 230000010354 integration Effects 0.000 claims abstract description 21
- 239000011159 matrix material Substances 0.000 claims abstract description 18
- 238000003708 edge detection Methods 0.000 claims abstract description 12
- 238000002474 experimental method Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 abstract description 3
- 231100000241 scar Toxicity 0.000 description 5
- 208000027418 Wounds and injury Diseases 0.000 description 4
- 238000005520 cutting process Methods 0.000 description 4
- 230000006378 damage Effects 0.000 description 4
- 208000014674 injury Diseases 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 241000196324 Embryophyta Species 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 235000009508 confectionery Nutrition 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20032—Median filtering
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of cane stalk recognition methods based on computer vision, belong to technical field of image processing.The method is by isolating the R component figure in black background sugarcane image rgb space, identify the position of R component figure middle period trace line, indirect identification cane stalk position, transverse direction sobel edge detection is carried out to sugarcane image after denoising simultaneously, when constructing rectangular integration operator, the width of matrix is that variable B takes [5, 15] odd number value in, and then the maximum value in circulation searching feature description vectors, pass through the minimum value of the distance between limitation stipes and eustipes part vector element, determine all stipes positions in sugarcane image, this method robustness is good, 95% can be reached by being experimentally confirmed stipes discrimination, its execution efficiency is high, individual sugarcane image recognition used time is less than 0.5s, the present invention can identify all stipes in sugarcane image by single treatment, it is simple and quick accurate.
Description
Technical field
The present invention relates to a kind of cane stalk recognition methods based on computer vision, belong to technical field of image processing.
Background technique
China is sugarcane production big country, the world, and the development of cane planting industry directly influences the livelihood of the tens million of sugarcane growers in China
With the development of sugar industry.Mostly realize to each sugarcane production in the world cane planting machinery to a certain extent, existing kind
Plant machine is mostly to cut kind of a formula sugarcane planting machine in real time, is all not carried out the bud injury preventing of planting process, sugarcane kind dosage in planting process
Greatly, due to being unable to bud injury preventing, effective sugarcane kind is few, and the development of the seedling in sugarcane field is irregular, eventually leads to that sugarcane yield is low, production cost
It is high.
Pre-cut kind formula planting machine is adapted to the plantation demand of the hilly grounds such as Guangxi, but pre-cut kind formula planting machine, needs pre-
Sugarcane is first cut into the sugarcane section containing simple bud, this work at present cuts kind by manually cutting kind and cutting the fixed length of kind of machine, manually cuts kind
It can all be occurred by hurting bud phenomenon with cutting kind of a machine fixed length and cutting to plant.Sugarcane based on machine vision cuts during kind of a machine can be realized and cut kind
Bud injury preventing function, need with machine vision technique identification cane stalk position, stipes information is sent to control system,
Sugarcane is cut by control system control cutter again.It is wherein core skill that bud injury preventing cuts kind of machine using Machine Vision Recognition stipes
Art, at present both at home and abroad this field research also in the exploratory stage.Close research has foreign countries Iran Moshashai K using ash
The method for spending carrying out image threshold segmentation has done Primary Study to cane stalk identification.The country has that yellow also it is equal using sugarcane image local
Value tag identifies cane stalk.Shi Changyou utilizes stipes adjacent margins line feature, has done sugarcane based on graph line detection algorithm
Stipes Study of recognition.
But existing recognition methods poor robustness, identification accuracy and efficiency are low.When sugarcane is because standing time is too long, surface is special
When sign changes, the detection method based on local mean value can not detect cane stalk.Edge line is miscellaneous at cane stalk
Disorderly without chapter, the stipes recognition accuracy based on line detection method can be reduced.
Summary of the invention
In order to solve presently, there are existing recognition methods poor robustness, the low problem of identification accuracy and efficiency, this hair
It is bright to provide a kind of cane stalk recognition method based on computer vision, which comprises
(1) the acquisition background that sugarcane image is arranged is black, acquires black background sugarcane image;
(2) collected black background sugarcane image sugarcane area-of-interest is obtained;
(3) the R component figure in black background sugarcane image rgb space is isolated, identification R component figure middle period trace line is passed through
Position, indirect identification cane stalk position;
(4) the fuzzy denoising of intermediate value carries out transverse direction sobel edge detection to sugarcane image after denoising;Sobel edge detection is public
Formula is as follows:
Gx=[f (x+1, y-1)+2*f (x+1, y)+f (x+1, y+1)]-[f (x-1, y-1)+2*f (x-1, y)+f (x-1, y
+1)]
Wherein f (x, y) indicates that the gray value of sugarcane image (x, y) point, A are R component image;
(5) rectangular integration operator is constructed;Rectangular integration operator is a matrix, and the element value in matrix at x row y column is Ixy,
The width of the height of a height of sugarcane image of matrix, matrix is variable B, and B is odd number;The width of matrix is determined according to experiment;
(6) feature description vectors are calculated, feature description vectors are normalized into [0,255] section;Rectangular integration operator from
Left-to-right covering sugarcane image, the element value in feature description vectors is rectangular integration operator and sugarcane image coincides part
The sum of corresponding element product, seeks the piecewise function V of feature description vectors elementx:
Wherein P (x, y) is the pixel value at sobel edge-detected image coordinate (x, y), this pixel value represents gradient value
Size;VxIt is characterized the element value of description vectors, subscript x is the call number of feature description vectors;W and H is sugarcane image respectively
Width value and height value;
(7) maximum value in circulation searching feature description vectors passes through the distance between limitation stipes and eustipes part vector
The minimum value of element determines all stipes positions in sugarcane image;
(8) by stipes position mark on sugarcane original image, and the image behind output token cane stalk position.
Optionally, the width B value of the matrix is the odd number value in [5,15].
Optionally, work as IxyWhen the value of all 1, B takes 11, rectangular integration operator RxAre as follows:
Wherein x is the call number of rectangular integration operator element, and H is the height of sugarcane image.
Optionally, the maximum value in described (6) the circulation searching feature description vectors, by limitation the distance between stipes and
The minimum value of eustipes part vector element determines all stipes positions in sugarcane image, comprising:
1. the maximum value of element first in search characteristics description vectors;
2. judging whether maximum value is greater than 125;
3. maximum value if more than 125, then assert that this position is stipes position, this location information is stored;
If returning to all stipes positions searched 4. maximum value, less than 125, stipes search finishes;
5. each 200 elements in the front and back of maximum value element are assigned a value of 0;
6. repeat 1. -5. finished to search.
Optionally, whether it is that the restrictive condition of stipes is at judging characteristic description vectors greatest member value:
I, the feature description vectors element value at stipes position is greater than 125;
II, stipes position of the peak-peak position as this stipes area in stipes area multi-peak is chosen;
III, the distance between stipes and stipes are greater than 200 pixels.
Optionally, the structural element size of the fuzzy denoising of the intermediate value is 5 × 5.
Optionally, sweet using the collected black background sugarcane image of VS2015 software acquisition configured with opencv3.2
Sugarcane area-of-interest.
The application also provides a kind of sugarcane and cuts kind of a machine, and the sugarcane cuts kind of machine and carries out cane stalk knowledge using the above method
Not.
The application also provides application of the above method in cane planting field.
The application also provides above-mentioned sugarcane and cuts application of kind of the machine in cane planting field.
The medicine have the advantages that
Cane stalk recognition method based on computer vision provided by the present application, by isolating black background sugarcane figure
As the R component figure in rgb space, the position of R component figure middle period trace line is identified, indirect identification cane stalk position, while to going
Sugarcane image carries out transverse direction sobel edge detection after making an uproar, when constructing rectangular integration operator, the width of matrix take for variable B [5,
15] odd number value in, and then the maximum value in circulation searching feature description vectors pass through the distance between limitation stipes and stipes
The minimum value for locating vector element, determines all stipes positions in sugarcane image, this method robustness is good, is experimentally confirmed stipes
Discrimination can reach 95%, and execution efficiency is high, individual sugarcane image recognition used time is less than 0.5s, and the present invention is by primary place
Reason can identify all stipes in sugarcane image, simple and quick accurate.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for
For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings other
Attached drawing.
Fig. 1 is sugarcane original image.
Fig. 2 is sobel edge-detected image.
Fig. 3 is characterized description vectors visualization figure.
Fig. 4 is stipes recognition result figure.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention
Formula is described in further detail.
Embodiment one:
The present embodiment provides a kind of cane stalk recognition methods based on computer vision.
The present embodiment uses the VS2015 software configured with opencv3.2 to carry out collected black background sugarcane image
Processing, comprising:
One, sugarcane area-of-interest is obtained, carries out identification stipes operation for sugarcane area-of-interest, Fig. 1 is original sugarcane
Beginning image.
Two, the R component in sugarcane image rgb space is isolated;
R component figure can clearly give expression to the feature of sugarcane image, and the brightness of cane stalk area is larger, the brightness of the area Jing Jian compared with
It is weak.The leaf scar feature in stipes area is obvious, is black at leaf scar line, is high brightness at left and right sides of leaf scar line, leaf scar is straight around sugarcane
Diameter one week, and the axis near normal with sugarcane.Position of the present invention by identification leaf scar line, indirect identification cane stalk section
It sets.It is as follows to separate R component figure code.
std::vector<cv::Mat>rgbChannels(3);
split(roiImg,rgbChannels);
Three, the fuzzy denoising of intermediate value carries out transverse direction sobel edge detection to sugarcane image after denoising;
Edge detection graph seems gradient image, from can be seen that stipes field gradient value in gradient image than field gradient between stem
It is worth greatly more.Stipes area is shown as after image viewing highlighted edge, the edge that the area Jing Jian has brightness darker.Sobel is horizontal
It is as follows to edge detection formula
Gx=[f (x+1, y-1)+2*f (x+1, y)+f (x+1, y+1)]-[f (x-1, y-1)+2*f (x-1, y)+f (x-1, y
+1)]
Wherein f (x, y) indicates the gray value of sugarcane image (x, y) point.
Intermediate value is fuzzy as follows with sobel edge detection edge detection code, and the size of intermediate value fuzzy structure element is 5 × 5.
The input parameter introduction of Sobel function, the 4th parameter are the difference orders on the direction x, and the 5th parameter is the difference on the direction y
Sublevel number, the 4th parameter and the 5th parameter be respectively 1,0 show the present invention use sobel transverse edge detect, Fig. 2
For sobel edge-detected image.Program process is as follows
medianBlur(rgbChannels[2],rgbChannels2medianBlur,5);
Sobel(rgbChannels2medianBlur,rgbChannels2medianBlurgrad_x,CV_32F,1,0,
3,1,1,BORDER_DEFAULT);
Src=rgbChannels2medianBlurgrad_x.clone ();
Four, rectangular integration operator is constructed.Rectangular integration operator is a matrix, and the element value in matrix at x row y column is 1, square
The high H of a height of sugarcane image of battle array, the width of matrix are 11.Rectangular integration operator
Wherein x is the call number of rectangular integration operator element, and H is the height of sugarcane image.
Five, feature description vectors are calculated, feature description vectors are normalized into [0,255] section.Rectangular integration operator from
Left-to-right covering sugarcane image, the element value in feature description vectors is rectangular integration operator and sugarcane image coincides part
The sum of corresponding element product, seeks the piecewise function of feature description vectors element
Wherein P (x, y) is the pixel value at sobel edge-detected image coordinate (x, y), this pixel value represents gradient value
Size.VxIt is characterized the element value of description vectors, subscript x is the call number of feature description vectors.W and H is sugarcane image respectively
Width value and height value.
Six, the maximum value in circulation searching feature description vectors passes through the distance between limitation stipes and eustipes part vector
The minimum value of element determines all stipes positions in sugarcane image.Search process is, 1. first first in search characteristics description vectors
The maximum value of element.2. judging whether maximum value is greater than 125.3. maximum value if more than 125, then assert that this position is stipes position,
This location information is stored.If returning to all stipes positions searched 4. maximum value, less than 125, stipes search finishes.⑤
Each 200 elements in the front and back of maximum value element are assigned a value of 0.6. repeat 1. -5. finished to search.Algorithmic procedure is as follows:
Seven, by stipes position mark on sugarcane original image, and the image behind output token cane stalk position.
It eight, is the recognition effect for verifying proposition method of the present invention, experiment chooses 119 sugarcane images and carries out stipes identification,
Experimental result is as shown in table 1, and stipes discrimination can reach 95%, individual sugarcane image recognition used time is 0.539s.The present invention
Method is high-efficient, accuracy is high.
1 stipes discrimination of table and used time
Cane stalk recognition method based on computer vision provided by the present application, by isolating black background sugarcane figure
As the R component figure in rgb space, the position of R component figure middle period trace line is identified, indirect identification cane stalk position, while to going
Sugarcane image carries out transverse direction sobel edge detection after making an uproar, when constructing rectangular integration operator, the width of matrix take for variable B [5,
15] odd number value in, and then the maximum value in circulation searching feature description vectors pass through the distance between limitation stipes and stipes
The minimum value for locating vector element, determines all stipes positions in sugarcane image, this method robustness is good, is experimentally confirmed stipes
Discrimination can reach 95%, and execution efficiency is high, individual sugarcane image recognition used time is less than 0.5s, and the present invention is by primary place
Reason can identify all stipes in sugarcane image, simple and quick accurate;The application is by user-defined feature description vectors, very well
Describe the Gradient Features of image, highlight the feature of eustipes part, stipes identification it is more rapidly accurate;And the application method
It is not influenced by sugarcane color and diameter change, there is adaptivity, identify precise and high efficiency.
Part steps in the embodiment of the present invention, can use software realization, and corresponding software program can store can
In the storage medium of reading, such as CD or hard disk.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and
Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of cane stalk recognition method based on computer vision, which is characterized in that the described method includes:
(1) the acquisition background that sugarcane image is arranged is black, acquires black background sugarcane image;
(2) collected black background sugarcane image sugarcane area-of-interest is obtained;
(3) the R component figure in black background sugarcane image rgb space is isolated, by the position for identifying R component figure middle period trace line
It sets, indirect identification cane stalk position;
(4) the fuzzy denoising of intermediate value carries out transverse direction sobel edge detection to sugarcane image after denoising;Sobel edge detection formula is such as
Under:
Gx=[f (x+1, y-1)+2*f (x+1, y)+f (x+1, y+1)]-[f (x-1, y-1)+2*f (x-1, y)+f (x-1, y+1)]
Wherein f (x, y) indicates that the gray value of sugarcane image (x, y) point, A are R component image;
(5) rectangular integration operator is constructed;Rectangular integration operator is a matrix, and the element value in matrix at x row y column is Ixy, matrix
A height of sugarcane image height, the width of matrix is variable B, and B is odd number;The width of matrix is determined according to experiment;
(6) feature description vectors are calculated, feature description vectors are normalized into [0,255] section;Rectangular integration operator from a left side to
Right covering sugarcane image, the element value in feature description vectors is rectangular integration operator to coincide the corresponding of part with sugarcane image
The sum of element product, seeks the piecewise function V of feature description vectors elementx:
Wherein P (x, y) is the pixel value at sobel edge-detected image coordinate (x, y), this pixel value represents the big of gradient value
It is small;VxIt is characterized the element value of description vectors, subscript x is the call number of feature description vectors;W and H is sugarcane image respectively
Width value and height value;
(7) maximum value in circulation searching feature description vectors passes through the distance between limitation stipes and eustipes part vector element
Minimum value, determine all stipes positions in sugarcane image;
(8) by stipes position mark on sugarcane original image, and the image behind output token cane stalk position.
2. the method according to claim 1, wherein the width B value of the matrix is the odd number in [5,15]
Value.
3. according to the method described in claim 2, it is characterized in that, working as IxyWhen the value of all 1, B takes 11, rectangular integration operator
RxAre as follows:
Wherein x is the call number of rectangular integration operator element, and H is the height of sugarcane image.
4. according to the method described in claim 3, it is characterized in that, maximum in (6) the circulation searching feature description vectors
Value determines all stipes positions in sugarcane image by the minimum value of the distance between limitation stipes and eustipes part vector element,
Include:
1. the maximum value of element first in search characteristics description vectors;
2. judging whether maximum value is greater than 125;
3. maximum value if more than 125, then assert that this position is stipes position, this location information is stored;
If returning to all stipes positions searched 4. maximum value, less than 125, stipes search finishes;
5. each 200 elements in the front and back of maximum value element are assigned a value of 0;
6. repeat 1. -5. finished to search.
5. according to the method described in claim 4, it is characterized in that, whether being stem at judging characteristic description vectors greatest member value
The restrictive condition of section is:
I, the feature description vectors element value at stipes position is greater than 125;
II, stipes position of the peak-peak position as this stipes area in stipes area multi-peak is chosen;
III, the distance between stipes and stipes are greater than 200 pixels.
6. the method according to claim 1, wherein the structural element size of the fuzzy denoising of the intermediate value be 5 ×
5。
7. the method according to claim 1, wherein being obtained using the VS2015 software configured with opencv3.2
Collected black background sugarcane image sugarcane area-of-interest.
8. a kind of sugarcane cuts kind of a machine, which is characterized in that the sugarcane cut kind of machine using method as claimed in claim 1 to 7 into
The identification of row cane stalk.
9. application of the method as claimed in claim 1 to 7 in cane planting field.
10. sugarcane according to any one of claims 8 cuts application of kind of the machine in cane planting field.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910654639.4A CN110400350A (en) | 2019-07-19 | 2019-07-19 | A kind of cane stalk recognition method based on computer vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910654639.4A CN110400350A (en) | 2019-07-19 | 2019-07-19 | A kind of cane stalk recognition method based on computer vision |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110400350A true CN110400350A (en) | 2019-11-01 |
Family
ID=68324774
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910654639.4A Pending CN110400350A (en) | 2019-07-19 | 2019-07-19 | A kind of cane stalk recognition method based on computer vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110400350A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113723314A (en) * | 2021-09-01 | 2021-11-30 | 江南大学 | Sugarcane stem node identification method based on YOLOv3 algorithm |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102930247A (en) * | 2012-10-18 | 2013-02-13 | 广西大学 | Sugarcane stalk node recognition method based on computer vision |
CN105654099A (en) * | 2014-08-25 | 2016-06-08 | 崔胡晋 | Sugarcane segmentation and identification method based on improved vision |
CN108960100A (en) * | 2018-06-22 | 2018-12-07 | 广西大学 | A kind of recognition methods of the sugarcane sugarcane section based on image procossing |
-
2019
- 2019-07-19 CN CN201910654639.4A patent/CN110400350A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102930247A (en) * | 2012-10-18 | 2013-02-13 | 广西大学 | Sugarcane stalk node recognition method based on computer vision |
CN105654099A (en) * | 2014-08-25 | 2016-06-08 | 崔胡晋 | Sugarcane segmentation and identification method based on improved vision |
CN108960100A (en) * | 2018-06-22 | 2018-12-07 | 广西大学 | A kind of recognition methods of the sugarcane sugarcane section based on image procossing |
Non-Patent Citations (1)
Title |
---|
DEQIANG ZHOU等: "Research on Algorithm of Sugarcane Nodes Identification Based on Machine Vision", 《2019 NICOGRAPH INTERNATIONAL (NICOINT)》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113723314A (en) * | 2021-09-01 | 2021-11-30 | 江南大学 | Sugarcane stem node identification method based on YOLOv3 algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Liu et al. | A simple pooling-based design for real-time salient object detection | |
Zheng et al. | Research on tomato detection in natural environment based on RC-YOLOv4 | |
CN102799854B (en) | Pattern recognition device and image-recognizing method | |
CN104346772B (en) | Thumbnail production method and device | |
CN111259925B (en) | K-means clustering and width mutation algorithm-based field wheat spike counting method | |
CN110400322A (en) | Fruit point cloud segmentation method based on color and three-dimensional geometric information | |
CN109859212A (en) | A kind of unmanned plane image soybean crops row dividing method | |
Khedaskar et al. | A survey of image processing and identification techniques | |
CN107220647A (en) | Crop location of the core method and system under a kind of blade crossing condition | |
CN112990103A (en) | String mining secondary positioning method based on machine vision | |
Rahman et al. | On the real-time semantic segmentation of aphid clusters in the wild | |
CN109670516A (en) | A kind of image characteristic extracting method, device, equipment and readable storage medium storing program for executing | |
CN110400350A (en) | A kind of cane stalk recognition method based on computer vision | |
Suo et al. | Casm-amfmnet: a network based on coordinate attention shuffle mechanism and asymmetric multi-scale fusion module for classification of grape leaf diseases | |
CN105354570A (en) | Method and system for precisely locating left and right boundaries of license plate | |
CN110378862A (en) | A kind of raising transmission line of electricity breaks the data enhancement methods of target identification accuracy outside | |
CN107578029B (en) | Computer-aided picture authentication method and device, electronic equipment and storage medium | |
CN107833227A (en) | A kind of method for drafting and its system of circular clipping region | |
CN116188855A (en) | Multi-scale plant disease identification method, device, storage medium and apparatus | |
Wu et al. | An improved YOLOv7 network using RGB-D multi-modal feature fusion for tea shoots detection | |
Wilms et al. | Localizing small apples in complex apple orchard environments | |
Zhu et al. | Exploring soybean flower and pod variation patterns during reproductive period based on fusion deep learning | |
Leite et al. | PhenoVis–A tool for visual phenological analysis of digital camera images using chronological percentage maps | |
CN115457581A (en) | Table extraction method and device and computer equipment | |
Cheng et al. | CACFTNet: A Hybrid Cov-Attention and Cross-Layer Fusion Transformer Network for Hyperspectral Image Classification |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191101 |
|
RJ01 | Rejection of invention patent application after publication |