CN111695560A - Method for actively positioning and focusing crop diseases and insect pests based on convolutional neural network - Google Patents
Method for actively positioning and focusing crop diseases and insect pests based on convolutional neural network Download PDFInfo
- Publication number
- CN111695560A CN111695560A CN202010397400.6A CN202010397400A CN111695560A CN 111695560 A CN111695560 A CN 111695560A CN 202010397400 A CN202010397400 A CN 202010397400A CN 111695560 A CN111695560 A CN 111695560A
- Authority
- CN
- China
- Prior art keywords
- picture
- image
- focusing
- disease
- characteristic
- 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
- 241000607479 Yersinia pestis Species 0.000 title claims abstract description 58
- 201000010099 disease Diseases 0.000 title claims abstract description 51
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 title claims abstract description 51
- 241000238631 Hexapoda Species 0.000 title claims abstract description 35
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 26
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000005516 engineering process Methods 0.000 claims abstract description 17
- 230000008569 process Effects 0.000 claims abstract description 9
- 230000004044 response Effects 0.000 claims abstract description 7
- 238000000605 extraction Methods 0.000 claims description 20
- 239000011159 matrix material Substances 0.000 claims description 18
- 239000013598 vector Substances 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 12
- 238000011176 pooling Methods 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 230000004913 activation Effects 0.000 claims description 3
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000011478 gradient descent method Methods 0.000 claims description 3
- 230000002708 enhancing effect Effects 0.000 claims description 2
- 238000011160 research Methods 0.000 abstract description 3
- 230000003313 weakening effect Effects 0.000 abstract 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
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
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- 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
-
- 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
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- 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
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a method for actively positioning and focusing crop diseases and insect pests based on a convolutional neural network, which belongs to the technical field of positioning and focusing of crop disease and insect pest parts, and aims to combine the latest convolutional neural network technology, obtain characteristic parameters of the disease and insect pest parts through the research of a large number of disease and insect pest part images, form a disease and insect pest part characteristic algorithm, judge the disease and insect pest parts only by substituting the image parameters into the algorithm, thereby automatically positioning and focusing the disease and insect pest parts in the photographing process, marking the disease and insect pest parts, weakening background interference and improving the accuracy and efficiency of disease and insect pest identification. The convolutional neural network technology is adopted, so that the focusing response speed is high; through the identification and classification of each pixel of the image, a plurality of pest and disease damage parts on one image can be identified simultaneously, so that the identification is more comprehensive; the automatic positioning and focusing self-learning can be helped by manually selecting and adjusting the positioning each time, and the characteristic parameters are updated, so that the positioning and focusing are more and more accurate.
Description
Technical Field
The invention relates to a method for actively positioning and focusing crop diseases and insect pests based on a convolutional neural network, and belongs to the technical field of positioning and focusing of crop disease and insect pest parts.
Background
At present, a great number of recognition technologies exist in the field of pest and disease recognition, and the image recognition methods adopted by the recognition technologies comprise an object image comparison recognition mode, an SVM statistical vector machine method and a convolutional neural network technology, the first two image recognition technologies are relatively lagged behind, and image recognition using convolutional neural network technology is also recognition based on the entire image, can accurately identify plant diseases and insect pests in a relatively perfect identification environment, reduces the identification rate when being put into an actual agricultural planting environment due to the interference of a complex and changeable field natural environment, in order to ensure that the complex image background is processed and identified after the image is uploaded according to the identification rate, the processing pressure at the server end is increased, the configuration requirement of image recognition on a hardware terminal is improved, and the recognition response time is greatly prolonged, so that the recognition efficiency is reduced.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a method for actively positioning and focusing crop diseases and insect pests based on a convolutional neural network.
The technical scheme for solving the technical problems is as follows:
a method for actively positioning and focusing crop diseases and insect pests based on a convolutional neural network comprises the following steps,
firstly, preprocessing an image when a user takes a picture, analyzing and processing an image picture by using a convolutional neural network technology, segmenting the image, and extracting the characteristics of each pixel;
secondly, according to the gray level change of the image, counting the contrast in the characteristics through gray level difference;
let (x, y) be a point in the image, and the gray scale difference value between the point and other points (x +. DELTA.x, y +. DELTA.y) with only a small distance is,
g△(x,y)＝g(x,y)-g(x+△x,y+△y)
in the formula, g delta is gray difference, delta X is a coordinate point X of a near point, and delta Y is a coordinate point Y of the near point;
setting all possible values of the gray difference to be m levels, moving the point (x, y) on the whole image, and accumulating the times of removing each numerical value of g delta (x, y), thereby making a histogram of g delta (x, y);
the histogram can know that g delta (x, y) takes the probability p delta (i), i takes the value between 1 and m, when the probability p delta (i) of a smaller i value is larger, the texture is rough, and when the values of p delta (i) are closer, namely the probability step is flatter, the texture is thinner.
A formula for calculating the contrast ratio is shown,
con＝∑i∧2p△(i)
the formula for calculating the direction of the angle,
ASM＝CON＝∑i[p△(i)]∧2
the entropy of the signal is,
ENT＝-∑ip△(i)lgp△(i)
the average value of the values is calculated,
MEAN＝1/m∑ip△(i)
thirdly, extracting the characteristics of the original color, and respectively extracting three color characteristics of a first-order matrix, a second-order matrix and a third-order matrix under a blue channel of the color image;
step four, extracting the characteristics of the texture characteristics, setting a plurality of texture parameters and extracting characteristic vectors;
step five, constructing parameters of roundness, rectangularity, eccentricity, sphericity ratio, compactness, breadth and inscribed circle radius as shape recognition characteristic vectors through shape characteristic extraction;
and step six, finally, integrating the characteristics, substituting all characteristic parameters into an algorithm to perform characteristic calculation, marking the pest and disease damage parts, and enabling the average response of the whole marking process to be within 2 s.
Preferably, the step of preprocessing the image when the user takes a picture in the first step is that firstly, the contrast of the whole picture is enhanced by changing the dynamic range of the grey value of the picture, then the picture is divided, picture features are extracted in different areas, then the features in each area are substituted into an insect pest feature algorithm for calculation, and finally, the calculation result is returned, and at this time, the area which accords with the features on the picture in the user taking the picture area is marked by a square frame.
Preferably, in the first step, the analysis processing of the image picture by using the convolutional neural network technology is to divide the image into small area pictures of one block according to a specified size; and then, extracting the characteristics of each region, substituting the characteristic parameters extracted from each region into a pest characteristic algorithm for calculation, returning the result, and marking the region by using a square box.
Preferably, the area marked by the box can be operated by self, the selected box can be deselected by double clicking, and the position can be selected by clicking any position of the picture.
Preferably, the feature extraction step in the first step is,
a, performing primary feature extraction on a picture by using a convolutional layer, decomposing the picture during extraction, then defining a convolution kernel for extracting a certain feature from the picture, multiplying the convolution kernel by a corresponding bit of a digital matrix and adding to obtain an output result of the convolutional layer;
b, inputting the output result of the convolutional layer into a pooling layer, wherein the pooling layer is used for reducing the dimension of the feature vector output by the convolutional layer, reducing the over-fitting phenomenon, only retaining the most useful picture information and reducing the transmission of noise;
c, after the extraction of the features of the convolution layer and the pooling layer is completed, generating a separator with the number equal to the required classes by using a full connection layer, cutting the tensor output by the pooling layer into a plurality of vectors again, multiplying the vectors by a weight matrix, adding an offset, and then optimizing the parameters by using a gradient descent method for the used ReLU activation function;
and d, obtaining characteristic parameters including the characteristics of the pest and disease damage parts such as color, texture, shape and the like through the process, repeatedly training for many times, continuously optimizing the characteristic parameters to only achieve the optimal effect, and obtaining a final pest and disease damage part characteristic matrix and a characteristic matching algorithm.
Compared with the prior art, the invention has the beneficial effects that: the invention aims to combine the latest convolutional neural network technology, obtain the characteristic parameters of the disease and insect pest part through the research of a large number of disease and insect pest part images, form a disease and insect pest part characteristic algorithm, judge the disease and insect pest part only by substituting the image parameters into the algorithm, automatically position and focus the disease and insect pest part before uploading the image, mark the disease and insect pest part in the photographing process, weaken background interference and improve the disease and insect pest identification accuracy and identification efficiency; by automatically positioning and focusing the pest and disease damage part, background interference is weakened, and the identification accuracy is improved; the convolutional neural network technology is adopted, so that the focusing response speed is high, and the recognition speed is high; through the identification and classification of each pixel of the image, a plurality of pest and disease damage parts on one image can be identified simultaneously, so that the identification is more comprehensive; the method has the advantages that the marking adjustment can be manually carried out on users with incomplete marking or wrong marking of the plant diseases and insect pests, positions can be selected or can be cancelled, meanwhile, the manually marked pictures can be transmitted to the feature sample library, the sample library carries out feature extraction on the manually marked parts through training and learning, original feature parameters are perfected through feature comparison, and therefore the more rigorous the feature matching algorithm is, and the automatic positioning focusing is more and more accurate.
Drawings
FIG. 1 is a flow chart of the operation of the present invention.
Fig. 2 is a working principle diagram of the present invention.
FIG. 3 is a graph of signature extraction according to the present invention.
FIG. 4 is a selected map of the marker diseases of the present invention.
Detailed Description
The principles and features of this invention are described below in conjunction with the following drawings, which are set forth by way of illustration only and are not intended to limit the scope of the invention.
As shown in fig. 1, a method for actively positioning and focusing crop pests based on a convolutional neural network comprises the following steps,
firstly, preprocessing an image when a user takes a picture, analyzing and processing an image picture by using a convolutional neural network technology, segmenting the image, and extracting the characteristics of each pixel instead of the characteristics of the whole image;
secondly, according to the gray level change of the image, counting the contrast in the characteristics through gray level difference;
let (x, y) be a point in the image, and the gray scale difference value between the point and other points (x +. DELTA.x, y +. DELTA.y) with only a small distance is,
g△(x,y)＝g(x,y)-g(x+△x,y+△y)
in the formula, g delta is gray difference, delta X is a coordinate point X of a near point, and delta Y is a coordinate point Y of the near point;
setting all possible values of the gray difference to be m levels, moving the point (x, y) on the whole image, and accumulating the times of removing each numerical value of g delta (x, y), thereby making a histogram of g delta (x, y);
the histogram can know that g delta (x, y) takes the probability p delta (i), i takes the value between 1 and m, when the probability p delta (i) of a smaller i value is larger, the texture is rough, and when the values of p delta (i) are closer, namely the probability step is flatter, the texture is thinner.
A formula for calculating the contrast ratio is shown,
con＝∑i∧2p△(i)
the formula for calculating the direction of the angle,
ASM＝CON＝∑i[p△(i)]∧2
the entropy of the signal is,
ENT＝-∑ip△(i)lgp△(i)
the average value of the values is calculated,
MEAN＝1/m∑ip△(i)
thirdly, extracting the characteristics of the original color, and respectively extracting three color characteristics of a first-order matrix, a second-order matrix and a third-order matrix under a blue channel of the color image;
step four, extracting the characteristics of the texture characteristics, setting a plurality of texture parameters and extracting characteristic vectors;
step five, constructing parameters of roundness, rectangularity, eccentricity, sphericity ratio, compactness, breadth and inscribed circle radius as shape recognition characteristic vectors through shape characteristic extraction;
and step six, finally, integrating the characteristics, substituting all characteristic parameters into an algorithm to perform characteristic calculation, marking the pest and disease damage parts, and enabling the average response of the whole marking process to be within 2 s.
And the step of preprocessing the image when the user takes a picture in the first step comprises the steps of firstly enhancing the overall contrast of the picture by changing the dynamic range of the gray value of the picture, then segmenting the picture, extracting the picture characteristics in regions, then substituting the characteristics in each region into an insect disease characteristic algorithm for calculation, and finally returning the calculation result, wherein the regions which accord with the characteristics on the picture in the user taking region are marked by boxes.
In the first step, the image picture is analyzed and processed by using the convolutional neural network technology, namely, the image is divided into small area graphs of one block according to the specified size, as shown in fig. 3; and then, extracting the characteristics of each region, substituting the characteristic parameters extracted from each region into a pest characteristic algorithm for calculation, returning the result, and marking the region by using a square box.
The area marked by the square frame can also be operated by self, the double-click on the selected square frame can cancel the selection, and the single click on any position of the picture can select the position, as shown in fig. 4.
As shown in fig. 2, the feature extraction step in the first step is,
a, performing primary feature extraction on a picture by using a convolutional layer, decomposing the picture during extraction, then defining a convolution kernel for extracting a certain feature from the picture, multiplying the convolution kernel by a corresponding bit of a digital matrix and adding to obtain an output result of the convolutional layer;
b, inputting the output result of the convolutional layer into a pooling layer, wherein the pooling layer is used for reducing the dimension of the feature vector output by the convolutional layer, reducing the over-fitting phenomenon, only retaining the most useful picture information and reducing the transmission of noise;
c, after the extraction of the features of the convolution layer and the pooling layer is completed, generating a separator with the number equal to the required classes by using a full connection layer, cutting the tensor output by the pooling layer into a plurality of vectors again, multiplying the vectors by a weight matrix, adding an offset, and then optimizing the parameters by using a gradient descent method for the used ReLU activation function;
and d, obtaining characteristic parameters including the characteristics of the pest and disease damage parts such as color, texture, shape and the like through the process, repeatedly training for many times, continuously optimizing the characteristic parameters to only achieve the optimal effect, and obtaining a final pest and disease damage part characteristic matrix and a characteristic matching algorithm.
The invention aims to combine the latest convolutional neural network technology, obtain the characteristic parameters of the disease and insect pest part through the research of a large number of disease and insect pest part images, form a disease and insect pest part characteristic algorithm, judge the disease and insect pest part only by substituting the image parameters into the algorithm, automatically position and focus the disease and insect pest part before uploading the image, mark the disease and insect pest part in the photographing process, weaken background interference and improve the disease and insect pest identification accuracy and identification efficiency; by automatically positioning and focusing the pest and disease damage part, background interference is weakened, and the identification accuracy is improved; the convolutional neural network technology is adopted, so that the focusing response speed is high, and the recognition speed is high; through the identification and classification of each pixel of the image, a plurality of pest and disease damage parts on one image can be identified simultaneously, so that the identification is more comprehensive; the automatic positioning focusing and manual selection positioning are matched for use, hundreds of accurate positioning of pest and disease damage parts is guaranteed, meanwhile, manually marked pictures can be transmitted to a sample library, the sample library performs feature extraction on manually marked parts through training and learning, original feature parameters are perfected through feature comparison, and therefore the feature matching algorithm is more accurate originally.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (5)
1. A method for actively positioning and focusing crop diseases and insect pests based on a convolutional neural network is characterized by comprising the following steps: comprises the following steps of (a) carrying out,
firstly, preprocessing an image when a user takes a picture, analyzing and processing an image picture by using a convolutional neural network technology, segmenting the image, and extracting the characteristics of each pixel;
secondly, according to the gray level change of the image, counting the contrast in the characteristics through gray level difference;
let (x, y) be a point in the image, and the gray scale difference value between the point and other points (x +. DELTA.x, y +. DELTA.y) with only a small distance is,
g△(x,y)＝g(x,y)-g(x+△x,y+△y)
in the formula, g delta is gray difference, delta X is a coordinate point X of a near point, and delta Y is a coordinate point Y of the near point;
setting all possible values of the gray difference to be m levels, moving the point (x, y) on the whole image, and accumulating the times of removing each numerical value of g delta (x, y), thereby making a histogram of g delta (x, y);
the histogram can know that g delta (x, y) takes the probability p delta (i) between 1 and m, when the probability p delta (i) of a smaller i value is larger, the texture is rough, when the values of p delta (i) are closer, namely the probability step is flatter, the texture is thinner,
a formula for calculating the contrast ratio is shown,
con＝∑i∧2p△(i)
the formula for calculating the direction of the angle,
ASM＝CON＝∑i[p△(i)]∧2
the entropy of the signal is,
ENT＝-∑ip△(i)lgp△(i)
the average value of the values is calculated,
MEAN＝1/m∑ip△(i)
thirdly, extracting the characteristics of the original color, and respectively extracting three color characteristics of a first-order matrix, a second-order matrix and a third-order matrix under a blue channel of the color image;
step four, extracting the characteristics of the texture characteristics, setting a plurality of texture parameters and extracting characteristic vectors;
step five, constructing parameters of roundness, rectangularity, eccentricity, sphericity ratio, compactness, breadth and inscribed circle radius as shape recognition characteristic vectors through shape characteristic extraction;
step six, finally, integrating the characteristics, substituting all characteristic parameters into an algorithm to perform characteristic calculation, and marking the pest and disease damage parts, wherein the average response of the whole marking process is within 2 s;
and seventhly, automatic positioning and focusing can be continuously self-learned, the marking adjustment can be manually carried out on a user with incomplete marking or wrong marking of the pest position, the position can be selected or the selection can be cancelled, meanwhile, the manually marked picture can be transmitted to a feature sample library, the sample library carries out feature extraction on the manually marked part through training and learning, original feature parameters are perfected through feature comparison, and therefore the more rigorous the feature matching algorithm is, and the automatic positioning and focusing are more and more accurate.
2. The active crop pest positioning and focusing method based on the convolutional neural network as claimed in claim 1, which is characterized in that: and the step of preprocessing the image when the user takes a picture in the first step comprises the steps of firstly enhancing the overall contrast of the picture by changing the dynamic range of the gray value of the picture, then segmenting the picture, extracting the picture characteristics in regions, then substituting the characteristics in each region into an insect disease characteristic algorithm for calculation, and finally returning the calculation result, wherein the regions which accord with the characteristics on the picture in the user taking region are marked by boxes.
3. The active crop pest positioning and focusing method based on the convolutional neural network as claimed in claim 1, which is characterized in that: in the first step, the image picture is analyzed and processed by using the convolutional neural network technology, namely, the image is divided into small area pictures of one block according to the specified size; and then, extracting the characteristics of each region, substituting the characteristic parameters extracted from each region into a pest characteristic algorithm for calculation, returning the result, and marking the region by using a square box.
4. The active crop pest positioning and focusing method based on the convolutional neural network as claimed in claim 3, wherein: the area marked by the square frame can be operated by self, double-clicking the selected square frame can cancel the selection, clicking any position of the picture can select the position, meanwhile, the manually marked picture can be transmitted to a sample library, the sample library carries out feature extraction on the manually marked part through training and learning, and the original feature parameters are perfected through feature comparison.
5. The active crop pest positioning and focusing method based on the convolutional neural network as claimed in claim 1, which is characterized in that: the characteristic extraction step in the first step is that,
a, performing primary feature extraction on a picture by using a convolutional layer, decomposing the picture during extraction, then defining a convolution kernel for extracting a certain feature from the picture, multiplying the convolution kernel by a corresponding bit of a digital matrix and adding to obtain an output result of the convolutional layer;
b, inputting the output result of the convolutional layer into a pooling layer, wherein the pooling layer is used for reducing the dimension of the feature vector output by the convolutional layer, reducing the over-fitting phenomenon, only retaining the most useful picture information and reducing the transmission of noise;
c, after the extraction of the features of the convolution layer and the pooling layer is completed, generating a separator with the number equal to the required classes by using a full connection layer, cutting the tensor output by the pooling layer into a plurality of vectors again, multiplying the vectors by a weight matrix, adding an offset, and then optimizing the parameters by using a gradient descent method for the used ReLU activation function;
and d, obtaining characteristic parameters including the characteristics of the pest and disease damage parts such as color, texture, shape and the like through the process, repeatedly training for many times, continuously optimizing the characteristic parameters to only achieve the optimal effect, and obtaining a final pest and disease damage part characteristic matrix and a characteristic matching algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010397400.6A CN111695560A (en) | 2020-05-12 | 2020-05-12 | Method for actively positioning and focusing crop diseases and insect pests based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010397400.6A CN111695560A (en) | 2020-05-12 | 2020-05-12 | Method for actively positioning and focusing crop diseases and insect pests based on convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111695560A true CN111695560A (en) | 2020-09-22 |
Family
ID=72477654
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010397400.6A Pending CN111695560A (en) | 2020-05-12 | 2020-05-12 | Method for actively positioning and focusing crop diseases and insect pests based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111695560A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114078125A (en) * | 2021-11-29 | 2022-02-22 | 开封市人民医院 | Vertebra image processing method based on nuclear magnetic image |
| TWI769724B (en) * | 2021-03-04 | 2022-07-01 | 鴻海精密工業股份有限公司 | Image feature extraction method and image feature extraction device, electronic device and storage media |
| CN116579751A (en) * | 2023-07-14 | 2023-08-11 | 南京信息工程大学 | Crop detection data processing method and system |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101901336A (en) * | 2010-06-11 | 2010-12-01 | 哈尔滨工程大学 | Fingerprint and finger vein bimodal recognition decision level fusion method |
| CN102842032A (en) * | 2012-07-18 | 2012-12-26 | 郑州金惠计算机系统工程有限公司 | Method for recognizing pornography images on mobile Internet based on multi-mode combinational strategy |
| CN103514459A (en) * | 2013-10-11 | 2014-01-15 | 中国科学院合肥物质科学研究院 | Method and system for identifying crop diseases and pests based on Android mobile phone platform |
| CN105787446A (en) * | 2016-02-24 | 2016-07-20 | 上海劲牛信息技术有限公司 | Smart agricultural insect disease remote automatic diagnosis system |
| CN107067043A (en) * | 2017-05-25 | 2017-08-18 | 哈尔滨工业大学 | A kind of diseases and pests of agronomic crop detection method |
| CN108510544A (en) * | 2018-03-30 | 2018-09-07 | 大连理工大学 | A kind of striation localization method of feature based cluster |
| CN110517311A (en) * | 2019-08-30 | 2019-11-29 | 北京麦飞科技有限公司 | Pest and disease monitoring method based on leaf spot lesion area |
-
2020
- 2020-05-12 CN CN202010397400.6A patent/CN111695560A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101901336A (en) * | 2010-06-11 | 2010-12-01 | 哈尔滨工程大学 | Fingerprint and finger vein bimodal recognition decision level fusion method |
| CN102842032A (en) * | 2012-07-18 | 2012-12-26 | 郑州金惠计算机系统工程有限公司 | Method for recognizing pornography images on mobile Internet based on multi-mode combinational strategy |
| CN103514459A (en) * | 2013-10-11 | 2014-01-15 | 中国科学院合肥物质科学研究院 | Method and system for identifying crop diseases and pests based on Android mobile phone platform |
| CN105787446A (en) * | 2016-02-24 | 2016-07-20 | 上海劲牛信息技术有限公司 | Smart agricultural insect disease remote automatic diagnosis system |
| CN107067043A (en) * | 2017-05-25 | 2017-08-18 | 哈尔滨工业大学 | A kind of diseases and pests of agronomic crop detection method |
| CN108510544A (en) * | 2018-03-30 | 2018-09-07 | 大连理工大学 | A kind of striation localization method of feature based cluster |
| CN110517311A (en) * | 2019-08-30 | 2019-11-29 | 北京麦飞科技有限公司 | Pest and disease monitoring method based on leaf spot lesion area |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI769724B (en) * | 2021-03-04 | 2022-07-01 | 鴻海精密工業股份有限公司 | Image feature extraction method and image feature extraction device, electronic device and storage media |
| CN114078125A (en) * | 2021-11-29 | 2022-02-22 | 开封市人民医院 | Vertebra image processing method based on nuclear magnetic image |
| CN116579751A (en) * | 2023-07-14 | 2023-08-11 | 南京信息工程大学 | Crop detection data processing method and system |
| CN116579751B (en) * | 2023-07-14 | 2023-09-08 | 南京信息工程大学 | Crop detection data processing method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113449594B (en) | Multilayer network combined remote sensing image ground semantic segmentation and area calculation method | |
| CN113160062B (en) | Infrared image target detection method, device, equipment and storage medium | |
| CN111783772A (en) | Grabbing detection method based on RP-ResNet network | |
| CN110070008A (en) | Bridge disease identification method adopting unmanned aerial vehicle image | |
| CN111695560A (en) | Method for actively positioning and focusing crop diseases and insect pests based on convolutional neural network | |
| CN111553200A (en) | Image detection and identification method and device | |
| CN110222767B (en) | Three-dimensional point cloud classification method based on nested neural network and grid map | |
| CN111860330A (en) | Apple leaf disease identification method based on multi-feature fusion and convolutional neural network | |
| CN113221956B (en) | Target identification method and device based on improved multi-scale depth model | |
| CN111178177A (en) | Cucumber disease identification method based on convolutional neural network | |
| CN110288033B (en) | Sugarcane top feature identification and positioning method based on convolutional neural network | |
| CN112464983A (en) | Small sample learning method for apple tree leaf disease image classification | |
| CN114170511A (en) | Pavement crack disease identification method based on Cascade RCNN | |
| CN116563205A (en) | Wheat spike counting detection method based on small target detection and improved YOLOv5 | |
| CN116977960A (en) | Rice seedling row detection method based on example segmentation | |
| CN109615610B (en) | Medical band-aid flaw detection method based on YOLO v2-tiny | |
| CN113344009B (en) | Light and small network self-adaptive tomato disease feature extraction method | |
| Zhong et al. | Identification and depth localization of clustered pod pepper based on improved Faster R-CNN | |
| CN113932712A (en) | Melon and fruit vegetable size measuring method based on depth camera and key points | |
| CN113449622A (en) | Image classification, identification and detection method for cotton plants and weeds | |
| CN116206208B (en) | Forestry plant diseases and insect pests rapid analysis system based on artificial intelligence | |
| CN114882303A (en) | Livestock counting method based on frame filtering and target detection | |
| CN111144464B (en) | Fruit automatic identification method based on CNN-Kmeans algorithm | |
| CN113920391A (en) | Target counting method based on generated scale self-adaptive true value graph | |
| CN111476129A (en) | Soil impurity detection method based on deep learning |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200922 |