CN105866040B - Bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot - Google Patents
Bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot Download PDFInfo
- Publication number
- CN105866040B CN105866040B CN201610177651.7A CN201610177651A CN105866040B CN 105866040 B CN105866040 B CN 105866040B CN 201610177651 A CN201610177651 A CN 201610177651A CN 105866040 B CN105866040 B CN 105866040B
- Authority
- CN
- China
- Prior art keywords
- image
- characteristic
- scab
- profile plot
- wave bands
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001228 spectrum Methods 0.000 title claims abstract description 46
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000009467 reduction Effects 0.000 title claims abstract description 19
- 241001272684 Xanthomonas campestris pv. oryzae Species 0.000 title claims abstract description 11
- 206010039509 Scab Diseases 0.000 claims abstract description 36
- 241000209094 Oryza Species 0.000 claims abstract description 19
- 235000007164 Oryza sativa Nutrition 0.000 claims abstract description 19
- 238000001514 detection method Methods 0.000 claims abstract description 19
- 235000009566 rice Nutrition 0.000 claims abstract description 19
- 230000001580 bacterial effect Effects 0.000 claims abstract description 9
- 230000003595 spectral effect Effects 0.000 claims description 10
- 238000004611 spectroscopical analysis Methods 0.000 claims description 7
- 230000036541 health Effects 0.000 claims description 4
- 201000010099 disease Diseases 0.000 abstract description 5
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 abstract description 5
- 230000003902 lesion Effects 0.000 abstract description 3
- 239000000284 extract Substances 0.000 abstract 1
- 238000002310 reflectometry Methods 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 3
- 238000007689 inspection Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 208000035143 Bacterial infection Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 239000000575 pesticide Substances 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 238000005507 spraying Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
-
- G06T3/06—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N2021/8466—Investigation of vegetal material, e.g. leaves, plants, fruits
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Abstract
The present invention discloses a kind of bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot, belongs to hyperspectral information and extracts field.This method comprises the following steps: the spectrum dimension first using successive projection algorithm in high spectrum image selects characteristic wave bands image, then corresponding profile plot is established to characteristic wave bands image, calculate the grey value difference of rice leaf different parts profile plot, minimal features wave band needed for can be realized the detection of rice leaf bacterial leaf-blight scab is further picked out, thus the dimension of high spectrum image needed for reducing scab detection.The method of the present invention combines successive projection algorithm and survey line drawing method, and effective dimensionality reduction of high dimensional data may be implemented, and obtained characteristic image can accurately identify scab;In conjunction with image recognition, lesion area and degree of disease are accurately calculated, the detection of bacterial blight of rice scab is realized with a small amount of characteristic image, the complexity of detection model is reduced, is effectively shortened detection the time it takes.
Description
Technical field
The invention belongs to hyperspectral informations to extract field, be related to spectral data analysis, and in particular to one kind is based on profile plot
Bacterial blight of rice high-spectrum image dimensionality reduction method.
Background technique
High-spectrum seems a three-dimensional data cube (Image Cube), and two dimensional image records the space shape of sample
State information, the corresponding spectral band of each image.Hyper-spectral image technique combines sample spectrum information and image information,
Realize nutrient estimation and the Defect inspection of crop.After rice infects bacterial leaf-blight, the scab that blade occurs can be gradually expanded, disease
Feelings aggravate;In farmland management, need to carry out spraying for pesticide according to the weight of the state of an illness.State of an illness inspection is carried out using high spectrum image
The main problem surveyed is that the data dimension up to 512 obtained is tieed up, and there are many wave band as involved in detection, processing speed
Slowly, the spectral information of not all wave band and image information can disclose the upgrowth situation of crop, on the contrary, many data
It may be some inessential noises and cover important information.Therefore, in order to realize the state of an illness detection real-time quick place
Reason needs preferably, reject uncorrelated or non-thread to the data of acquisition to improve model prediction accuracy and simplified model
Property variable, obtains the calibration model that predictive ability is strong, robustness is good.
Mainly whether there is or not the null methods of information variable, interval offset minimum binary for characteristic wave bands selection method based on spectrum dimension
Method, genetic algorithm, successive projection algorithm etc..The working principle of successive projection algorithm is to find to contain minimum in spectrum matrix
The set of variables of the redundancy of degree makes the synteny between variable reach minimum.
The method of profile plot based on leaf image is analyzed, root by the different piece gray value to leaf image
The characteristic wave bands based on image dimension are chosen according to the gray difference degree under different-waveband.
Since there are many wave band that high spectrum image detection is related to, the characteristic wave bands extraction for relying solely on spectrum dimension has been not easy to obtain
To the optimal combination for describing Disease Characters.
Summary of the invention
In order to overcome the disadvantages and deficiencies of the prior art, the purpose of the present invention is to provide a kind of rice based on profile plot
Bacterial leaf-blight high-spectrum image dimensionality reduction method.Present invention combination successive projection algorithm and the available better feature of profile plot method
Band combination, dimension needed for reducing scab detection, simplifies detection model.
The purpose of the invention is achieved by the following technical solution:
A kind of bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot, includes the following steps:
Spectrum dimension first using successive projection algorithm in high spectrum image selects characteristic wave bands image, then to characteristic wave
Section image establishes corresponding profile plot, calculates the grey value difference of rice leaf different parts profile plot, further picks out energy
Minimal features wave band needed for enough realizing the detection of rice leaf bacterial leaf-blight scab, thus bloom needed for reducing scab detection
The dimension of spectrogram picture.
The bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot, specifically comprises the following steps:
Realization pretreatment to the spectroscopic data of n sample, the spectroscopic data of each sample are xi=(xi1,
xi2..., xi512)T, wherein i=1,2 ..., n;
To sample xi, m characteristic wave bands of spectral information application successive projection algorithms selection are tieed up to high spectrum image 512, it is special
Sign Band Set is T={ t1, t2..., tm, m < 512;
Choose sample xiCharacteristic wave bands tl(l=1,2 ..., m) corresponding characteristic image
In sample xiCharacteristic imageThe middle profile plot for drawing rice leaf;
Sample x is calculated according to profile plotiCharacteristic imageMiddle blade health position (H), disease
Spot position (L), shade position (S) average gray GH, GL, GS;
To sample xiCharacteristic imageCalculate GHWith GLDifference CHL;Work as CHL>=100, table
Bright this feature image can distinguish healthy position and scab position, choose tlInto the spy that can distinguish healthy position and scab position
Levy Band Set S1;
To sample xiCharacteristic imageCalculate GHWith GSDifference CHS;Work as CHS>=100, table
Bright this feature image can distinguish healthy position and shade position, choose tlInto the spy that can distinguish healthy position and shade position
Levy Band Set S2;
To sample xiCharacteristic imageCalculate GLWith GSDifference CLS;Work as CLS>=100, table
Bright this feature image can distinguish scab position and shade position, choose tl into the spy that can distinguish scab position and shade position
Levy Band Set S3;
Characteristic wave bands needed for detecting rice leaf bacterial leaf-blight scab are S=S1∪S2∪S3, S ∈ T.Needed for scab detection
Characteristic wave bands quantity further decrease, therefore, it can be achieved that scab detection high spectrum image dimensionality reduction.
To the spectroscopic data x of sampleiThe pretreatment of progress is determined by formula 1, if kth dimension spectral reflectivity is xik(k=
1,2 ..., 512):
Wherein, WjIndicate the weight obtained using least square method.
The algorithm of the m characteristic wave bands of selection, comprising the following steps:
Assuming that initial is iterative vectorized for xk(0), the variable number for needing to extract is N, and spectrum matrix column variable number is J
It is a, then:
(1) before the 1st iteration (n=1), any 1 column j of optional spectrum matrix is assigned to correction spectrum battle array jth column
xj, it is denoted as xk(0);
(2) the unknown set of the column vector being selected into not yet is denoted as s, wherein
(3) x is calculated separatelyjProjection to remaining column vector:
(4) remember k (n)=arg [max (| | Pxj| |), j ∈ s];
(5) x is enabledj=Pxj, j ∈ s;
(6) n=n+1, if n < N, returns to (2) step cycle calculations;The variable finally extracted: { xk(n)=0 ...,
N-1}。
The blade health position (H), scab position (L), shade position (S) average gray GH, GL, GSRespectively
It is determined by formula 2, formula 3, formula 4:
If healthy span access location length is lH, scab span access location length be lL, shade span access location length be lS, then
The difference CHLIt is determined by formula 5:
CHL=GH-GL(formula 5)
The difference CHSIt is determined by formula 6:
CHS=GH-GS(formula 6)
The difference CLSIt is determined by formula 7:
CLS=GL-GS(formula 7).
The present invention compared with the existing technology, have following advantages and effects
(1) method that the present invention analyzes high-spectral data has organically combined spectrum peacekeeping image dimension, to EO-1 hyperion
Data carry out 2 dimensionality reductions, compared to the method for carrying out characteristic wave bands selection is only tieed up from spectrum, can select less characteristic wave
Section.
(2) the survey line drawing method used in method of the invention is to carry out operation to the gray scale of image, and data volume is small, processing
Speed is fast.
(3) the method for the present invention combines successive projection algorithm and survey line drawing method, and the effective of high dimensional data may be implemented
Dimensionality reduction, obtained characteristic image can accurately identify scab;In conjunction with image recognition, lesion area and disease journey are accurately calculated
Degree.
(4) present invention combines the spectrum dimension and image dimension of high spectrum image, realizes the white leaf of rice with a small amount of characteristic image
The detection of blight scab reduces the complexity of detection model, is effectively shortened detection the time it takes.
Detailed description of the invention
Fig. 1 is the dimensionality reduction flow chart of the invention based on successive projection algorithm and survey line drawing method to high spectrum image.
Fig. 2 is the scab position of rice leaf infection bacterial leaf-blight and the curve of spectrum comparison diagram at healthy position.
Fig. 3 is the characteristic wave bands that the successive projection algorithm of rice leaf obtains.
Fig. 4 is the characteristic spectrum image of blade.
Fig. 5 is the profile plot of rice leaf.
Fig. 6 is the scab segmentation result figure of characteristic image after dimensionality reduction.
Specific embodiment
Present invention will now be described in further detail with reference to the embodiments and the accompanying drawings, but embodiments of the present invention are unlimited
In this.
Embodiment 1
It is of the invention based on successive projection algorithm and survey line drawing method to the dimensionality reduction flow chart of high spectrum image, see Fig. 1 institute
Show.Present invention is mainly applied to the analyses of the high-spectral data of susceptible rice leaf.After rice infects bacterial leaf-blight, blade occurs
Scab and healthy leaves all show difference on spectral reflectivity and image.From the point of view of spectral reflectivity, catch an illness blade and
The curve of spectrum of healthy leaves is as shown in Figure 2.According to the present invention, the spectral reflectivity of blade is handled and is extracted first spy
Levy wave band.
Using the curve of spectrum at the healthy position of multiple samples and scab position as object, smooth pre- place is carried out to spectroscopic data
Reason;The spectroscopic data of each sample is xi=(xi1, xi2..., xi512)T, wherein i=1,2 ..., n;To high spectrum image
512 dimension spectral informations realize the dimensionality reduction that high spectrum image is tieed up in spectrum, such as Fig. 3 using successive projection algorithm picks characteristic wave bands
It is shown.Have chosen m characteristic wave bands, T={ t1, t2..., tm, m < 512.
The corresponding characteristic image of characteristic wave bands that spectrum dimension is chosen is extracted, as shown in Figure 4;A line is defined on blade to pass through
Blade is worn, the gray value of line corresponding points is calculated, the corresponding profile plot of leaf image is obtained, as shown in Fig. 5;The horizontal axis of profile plot
For length, i.e., the point between starting point at a distance from, the longitudinal axis is the corresponding gray value of point.
Calculate blade xiCharacteristic image in the healthy position (H) of profile plot blade, scab position (L), shade position (S)
Average gray GH, GL, GS。
To sample xiCharacteristic wave bands tl(l=1,2 ..., m) corresponding characteristic imageMeter
Calculate GHWith GLDifference CHL.Work as CHL>=100, show that this feature image can distinguish healthy position and scab position, chooses tlInto
Enter to distinguish the characteristic wave bands set S at healthy position and scab position1。
To sample xiCharacteristic wave bands tl(l=1,2 ..., m) corresponding characteristic imageMeter
Calculate GHWith GSDifference CHS.Work as CHS>=100, show that this feature image can distinguish healthy position and shade position, chooses tlInto
Enter to distinguish the characteristic wave bands set S at healthy position and shade position2。
To sample xiCharacteristic wave bands tl(l=1,2 ..., m) corresponding characteristic imageMeter
Calculate GLWith GSDifference CLS.Work as CLS>=100, show that this feature image can distinguish scab position and shade position, chooses tlInto
Enter to distinguish the characteristic wave bands set S at scab position and shade position3。
Realizing that rice leaf bacterial leaf-blight scab detects required characteristic wave bands is S=S1∪S2∪S3, S ∈ T, using these
Characteristic wave bands realize the differentiation of blade health part, scab part and dash area, and the corresponding image of S is characteristic image.
Image segmentation is carried out using OTSU method in characteristic image, as shown in Figure 6;It, can be accurately according to segmentation result
Calculate lesion area and state of an illness grade.
The above embodiment is a preferred embodiment of the present invention, but embodiments of the present invention are not by above-described embodiment
Limitation, other any changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principles of the present invention,
It should be equivalent substitute mode, be included within the scope of the present invention.
Claims (4)
1. a kind of bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot, it is characterised in that include the following steps:
Spectrum dimension first using successive projection algorithm in high spectrum image selects characteristic wave bands image, then to characteristic wave bands figure
As establishing corresponding profile plot, the grey value difference of rice leaf different parts profile plot is calculated, further picking out can be real
Minimal features wave band needed for existing rice leaf bacterial leaf-blight scab detection, thus high-spectrum needed for reducing scab detection
The dimension of picture;
Specifically comprise the following steps:
The spectroscopic data of Y sample is pre-processed, the spectroscopic data of each sample is xi=(xi1, xi2..., xi512)T,
Wherein i=1,2 ..., Y;
To sample xi, m characteristic wave bands of spectral information application successive projection algorithms selection, characteristic wave bands are tieed up to high spectrum image 512
Collection is combined into T={ t1, t2..., tm, m < 512;
Choose sample xiCharacteristic wave bands tl(l=1,2 ..., m) corresponding characteristic image
In sample xiCharacteristic imageThe middle profile plot for drawing rice leaf;
Sample x is calculated according to profile plotiCharacteristic imageMiddle blade health position, scab position, yin
The average gray G at shadow positionH, GL, GS;
To sample xiCharacteristic imageCalculate GHWith GLDifference CHL;Work as CHL>=100, choose tlInto
Enter the characteristic wave bands set S for distinguishing healthy position and scab position1;
To sample xiCharacteristic imageCalculate GHWith GSDifference CHS;Work as CHS>=100, choose tlInto
Enter the characteristic wave bands set S for distinguishing healthy position and shade position2;
To sample xiCharacteristic imageCalculate GLWith GSDifference CLS;Work as CLS>=100, choose tlInto
Enter to distinguish the characteristic wave bands set S at scab position and shade position3;
Characteristic wave bands needed for detecting rice leaf bacterial leaf-blight scab are S=S1∪S2∪S3, S ∈ T.
2. the bacterial blight of rice high-spectrum image dimensionality reduction method according to claim 1 based on profile plot, feature exist
In:
The difference CHLIt is determined by formula 5:
CHL=GH-GL(formula 5).
3. the bacterial blight of rice high-spectrum image dimensionality reduction method according to claim 1 based on profile plot, feature exist
In:
The difference CHSIt is determined by formula 6:
CHS=GH-GS(formula 6).
4. the bacterial blight of rice high-spectrum image dimensionality reduction method according to claim 1 based on profile plot, feature exist
In:
The difference CLSIt is determined by formula 7:
CLS=GL-GS(formula 7).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610177651.7A CN105866040B (en) | 2016-03-25 | 2016-03-25 | Bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610177651.7A CN105866040B (en) | 2016-03-25 | 2016-03-25 | Bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105866040A CN105866040A (en) | 2016-08-17 |
CN105866040B true CN105866040B (en) | 2019-02-19 |
Family
ID=56625860
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610177651.7A Active CN105866040B (en) | 2016-03-25 | 2016-03-25 | Bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105866040B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110873786A (en) * | 2019-01-03 | 2020-03-10 | 中化现代农业有限公司 | Pear sensory quality evaluation system and evaluation method |
CN115131683B (en) * | 2022-08-25 | 2022-12-09 | 金乡县林业保护和发展服务中心(金乡县湿地保护中心、金乡县野生动植物保护中心、金乡县国有白洼林场) | Forestry information identification method based on high-resolution remote sensing image |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102706813A (en) * | 2012-05-31 | 2012-10-03 | 北京林业大学 | Poa pratensis variety identification method based on hyper-spectral image |
CN103134758A (en) * | 2013-01-29 | 2013-06-05 | 华南农业大学 | Rice leaf blast disease resistance identification grading method based on multi-scale hyperspectral image processing |
CN104008394A (en) * | 2014-05-20 | 2014-08-27 | 西安电子科技大学 | Semi-supervision hyperspectral data dimension descending method based on largest neighbor boundary principle |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8280144B2 (en) * | 2007-02-21 | 2012-10-02 | Goldfinch Solutions, Llc | System and method for analyzing material properties using hyperspectral imaging |
US8315472B2 (en) * | 2009-05-29 | 2012-11-20 | Raytheon Company | System and method for reducing dimensionality of hyperspectral images |
-
2016
- 2016-03-25 CN CN201610177651.7A patent/CN105866040B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102706813A (en) * | 2012-05-31 | 2012-10-03 | 北京林业大学 | Poa pratensis variety identification method based on hyper-spectral image |
CN103134758A (en) * | 2013-01-29 | 2013-06-05 | 华南农业大学 | Rice leaf blast disease resistance identification grading method based on multi-scale hyperspectral image processing |
CN104008394A (en) * | 2014-05-20 | 2014-08-27 | 西安电子科技大学 | Semi-supervision hyperspectral data dimension descending method based on largest neighbor boundary principle |
Non-Patent Citations (2)
Title |
---|
因子分析模型的高光谱数据降维方法;李娜 等;《中国图象图形学报》;20111130;第16卷(第11期);2030-2035 |
高光谱数据降维技术研究;王旭红等;《水土保持通报》;20061231;第26卷(第6期);89-91 |
Also Published As
Publication number | Publication date |
---|---|
CN105866040A (en) | 2016-08-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Zhang et al. | Branch detection for apple trees trained in fruiting wall architecture using depth features and Regions-Convolutional Neural Network (R-CNN) | |
Paulus et al. | Surface feature based classification of plant organs from 3D laserscanned point clouds for plant phenotyping | |
WO2017049994A1 (en) | Hyperspectral image corner detection method and system | |
CN109146948A (en) | The quantization of crop growing state phenotypic parameter and the correlation with yield analysis method of view-based access control model | |
CN106485740B (en) | A kind of multidate SAR image registration method of combination stable point and characteristic point | |
CN107240122A (en) | Video target tracking method based on space and time continuous correlation filtering | |
CN111784721B (en) | Ultrasonic endoscopic image intelligent segmentation and quantification method and system based on deep learning | |
CN103150577B (en) | Method for selecting high spectrum remotely sensed image self-adaptive wave band based on particle swarm optimization | |
CN108664838A (en) | Based on the monitoring scene pedestrian detection method end to end for improving RPN depth networks | |
CN106339674A (en) | Hyperspectral image classification method based on edge preservation and graph cut model | |
Zhang et al. | Extraction of tree crowns damaged by Dendrolimus tabulaeformis Tsai et Liu via spectral-spatial classification using UAV-based hyperspectral images | |
CN110390255A (en) | High-speed rail environmental change monitoring method based on various dimensions feature extraction | |
CN107145831B (en) | Hyperspectral image classification method based on vector probability diffusion and Markov random field | |
CN105866040B (en) | Bacterial blight of rice high-spectrum image dimensionality reduction method based on profile plot | |
CN111586605A (en) | KNN indoor target positioning method based on adjacent weighted self-adaptive k value | |
Li et al. | Detection of powdery mildew on strawberry leaves based on DAC-YOLOv4 model | |
CN109446962A (en) | Land cover interannual variance detection method and system | |
CN106407975B (en) | Multiple dimensioned layering object detection method based on space-optical spectrum structural constraint | |
Zhang et al. | Branch detection with apple trees trained in fruiting wall architecture using stereo vision and regions-convolutional neural network (R-CNN) | |
Bao et al. | Method for wheat ear counting based on frequency domain decomposition of MSVF-ISCT | |
CN109829459A (en) | Based on the vision positioning method for improving RANSAC | |
CN103679703B (en) | A kind of target in hyperspectral remotely sensed image dimension reduction method based on conformal geometric algebra | |
CN108921884A (en) | Based on the optics and SAR Image registration method, equipment and storage medium for improving SIFT | |
Farooque et al. | Swin transformer with multiscale 3D atrous convolution for hyperspectral image classification | |
CN116863341A (en) | Crop classification and identification method and system based on time sequence satellite remote sensing image |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |