CN108537267A - A kind of corncob automatic identifying method based on significance test - Google Patents
A kind of corncob automatic identifying method based on significance test Download PDFInfo
- Publication number
- CN108537267A CN108537267A CN201810295953.3A CN201810295953A CN108537267A CN 108537267 A CN108537267 A CN 108537267A CN 201810295953 A CN201810295953 A CN 201810295953A CN 108537267 A CN108537267 A CN 108537267A
- Authority
- CN
- China
- Prior art keywords
- corncob
- picture
- feature
- notable
- identifying method
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of corncob automatic identifying method based on significance test, belongs to field Motion parameters field.This method includes two stages：The corncob potential region recognition stage, it is vulnerable to area light saturated phenomenon caused by strong shadow is rung for picture, it is delustered using one kind and being saturated algorithm progress image preprocessing, significance test is carried out to the saturation picture that delusters, marking area is as the potential region of corncob in extraction picture；It misidentifies region and removes the stage, the textural characteristics for extracting each marking area carry out classification removal background area, obtain final corncob recognition result.The characteristics of present invention is vulnerable to illumination effect for field shooting picture, the selection saturation algorithm that delusters pre-processes picture, improves the accuracy of identification of corncob, while carrying out automatic identification to corncob based on significance test, it calculates simple and effective, and has many advantages, such as higher robustness.
Description
Technical field
The invention belongs to field Motion parameters fields, and in particular to a kind of corncob region of field shooting picture from
Dynamic recognition methods is carried on the back in corncob region from complexity using significance test using saturation algorithm removal light zone of saturation of delustering
It is accurately extracted in scape.
Background technology
Corn is to plant one of widest cereal crops in the world, and total cultivated area is only second to wheat and rice, and in
State is second-biggest-in-the-world maize production state.Corn is as the most commonly used non-staple food raw materials for production of purposes in cereal crops and herding
Industry feed is reprocessed the production for being also widely used in weaving, papermaking, bio-fuel and cosmetics using corn as raw material
Etc. in technologies.Research about application of the machine vision technique in agricultural technology starts from the latter stage seventies, and what is be substantially carried out is
The automatic identification of field target and classification, the detection of quality of agricultural product and classification etc..With computer software and hardware technology
Rapid development, machine vision technique has larger progress in application study agriculturally.Compared to traditional artificial observation
Mode, machine vision and image analysis technology based on image procossing have in Crop Information acquisition contain much information, quickly, it is high
The notable advantage such as precision is saving labour, and the subjective impact etc. when reducing artificial observation has huge potentiality, can
To help people to analyze the relationship between crop growth state and agricultural weather, corresponding field operation is carried out in time, is provided
Effective agricultural helps, to increase agricultural output.
Ear period is one of most important growth period during corn growth, it implies that corn starts to pollinate, this
Period determines the fertilization of plant success.Corncob is the Main Agronomic Characters of corn breeding, and can be used as jade
Rice reaches the typical characteristics of ear period.The definition traditional to corn ear period refers to the corn tassel blossom more than half region, when in
When this stage, corn obviously speeds the absorption of nitrogen and phosphorus.If the growth conditions to corn monitor in time, you can right
Corresponding agricultural operation, which is improved, to be achieved the effect that increase yield.There are some researchs to have been proven that the feature of corncob can shadow
Ring the yield of corn.Such as the smaller corncob of shape can make sunlight more effectively be irradiated to plant, so as to cause production
The increase of amount.Traditional corncob recognition methods is mainly by artificial observation, but this mode is of high cost and subjectivity is strong, when long
Between continuous observation can influence observation result and efficiency.It is studied although people have carried out part to the identification of corncob, this
A little researchs do not account for the influence of environmental factor primarily directed to zonule sample.In field crop, if can be in time
The information that corn reaches ear period is obtained, peasant can carry out corresponding field operation in time, and this avoid unnecessary artificial
Observation is observed difficult caused by being especially that of avoiding extreme weather.Moreover, automatic identification ear period is identification corn growing season
Important component.Therefore, automatic identification corn ear period is very necessary for modern agriculture.
Invention content
The purpose of the present invention is overcoming the above technology, a kind of corncob based on significance test is provided and is known automatically
Other method, can realize the automatic identification of field corn fringe, and the algorithm is simple, quickly and effectively, and have higher robustness
The advantages that.
Technical scheme of the present invention：
A kind of corncob automatic identifying method based on significance test, includes the following steps：
(1) the potential region recognition of corncob：
Field picture when S1, acquisition corn are in ear period is as experiment sample；
S2, the depth of view information based on simulated light zone of saturation carry out saturated process of delustering to field picture, obtain delustering full
And picture；
S3, brightness, the color characteristic that the saturation each pixel of picture of delustering is calculated using Itti significance test algorithms
And direction character, and three features are subjected to linear superposition, obtain the notable figure after Fusion Features；
S4, the pixel region that saliency value is high in notable figure is extracted using adaptive threshold fuzziness, obtains the potential area of corncob
Domain；
(2) misrecognition region removal：
S5, the picture feature for extracting each potential region of corncob；
S6, LSSVM disaggregated models are established, training sample is divided into corncob region and background area；
S7, it determines corncob region, removes background area, obtain the final recognition result of corncob.
The S2 the specific steps are：
1) the brightness v and saturation degree s for extracting pixel in the picture of field, establish image depth d between image slices vegetarian refreshments x
Linear model：
D (x)=θ0+θ1v(x)+θ2s(x)+ε(x) (1)
2) relationship that light saturation image I is obtained between the saturation image J that delusters based on atmospherical scattering model：
I (x)=J (x) t (x)+A (1-t (x))
T (x)=e-βd(x) (2)
Wherein, θ0、θ1、θ2For linear coefficient, it is random error to take 0.12,0.96 and -0.78, ε respectively, and it is 0 side to take mean value
The Gaussian function that difference is 0.05, A are picture luminance peak, and t is transmission function, and between taking 0.1 to 0.9, β is atmospheric scattering system
Number, takes 1；
3) saturation image expression formula of delustering is obtained：
The S3 carries out multiple feature channels and multiple dimensioned decomposition to input picture first, establishes 9 layers of gaussian pyramid,
Then 9 layers of gaussian pyramid are filtered to obtain the characteristic pattern of each feature, are obtained respectively finally by characteristic pattern fusion calculation
The notable figure that a pixel saliency value is constituted, the specific steps are：
4) 9 layers of gaussian pyramid is established by input picture, wherein the 0th layer is input picture, the 1 to 8th layer is through 5 respectively
× 5 Gaussian filter is filtered input picture and samples the image obtained, and size is respectively the 1/2,1/ of input picture
4,1/8,1/16,1/32,1/64,1/128,1/256；Then brightness spy is extracted respectively to each tomographic image of gaussian pyramid
Sign, red feature, green characteristic, blue characteristics, yellow characteristic, direction character form brightness pyramid, chromaticity gold
Word tower and direction character pyramid；
5) feature takes difference to form characteristic pattern between different scale：Respectively in pyramidal center scale c (the c ∈ of different characteristic
{ 2,3,4 }) work is poor between edge scale s (s=c+ δ, δ ∈ { 3,4 }), brightness figure, color characteristic figure and side is calculated
To characteristic pattern；
6) characteristic pattern of the characteristic pattern of different scale and different characteristic is merged, generates notable figure：Use normalization
Function is first normalized the characteristic pattern of different scale in each feature, forms the synthesis notable figure of a width this feature,
The notable figure of different characteristic is subjected to linear, additive again and takes mean value, obtains overall visual notable figure.
The S4 the specific steps are：If the pixel significance distribution of every width notable figure meets Gaussian Profile rule, extraction is aobvious
The mean value m and variance d for scheming each pixel are write, m ± 3d is calculated as notable figure conspicuousness segmentation threshold, extracts saliency value in notable figure
Higher pixel region carries out binary conversion treatment to original image and obtains the potential region of corncob.
Picture feature in the S5 includes grey level histogram featureAnd gray level co-occurrence matrixes
FeatureInclude L sample and Num feature；The grey level histogram feature include mean value,
Six standard deviation, smoothness, third moment, consistency, gray level entropy features；The gray level co-occurrence matrixes feature includes on four direction
Contrast, correlation, energy and homogeney textural characteristics, totally ten six features.
The experiment sample is the picture that corn is in ear period between 3 years, and wherein First Year randomly selects half picture conduct
Training sample.
Corncob automatic identifying method of the present invention based on significance test, is mainly utilized corncob relative to the back of the body
The conspicuousness of scape carries out automatic identification, while can using the saturation algorithm that delusters for the influence of illumination factor in actual environment
Picture light zone of saturation is effectively removed, the automatic identification precision of corncob is improved.Automatic monitoring of this method for corn, crop
Growth conditions judgement has great significance.
Description of the drawings
Fig. 1 is the overall flow figure of field corn fringe automatic identifying method of the present invention；
Specific implementation mode
With reference to specific attached drawing, the invention will be further described.
A kind of corncob automatic identifying method based on significance test of the present invention, it is therefore intended that in practical field
In the environment of, influence of the environmental factors such as illumination to recognition effect is effectively removed using a kind of algorithm, realizes the automatic of corncob
Identification：As shown in Figure 1, specifically including, steps are as follows：
(1) the potential region recognition of corncob：Picture collection system is placed on the field that corn is in ear period, is existed to corn
The picture of ear period is shot, and corn ear period picture is as experiment sample, the potential region of identification corncob between extraction 3 years.Specifically
Steps are as follows：
A, picture delusters saturation, and extraction corn field picture extracts the brightness v and saturation degree s of pixel, establishes image scape
Deep d is the same as the linear model between image slices vegetarian refreshments x：
D (x)=θ0+θ1v(x)+θ2s(x)+ε(x) (1)
The relationship for obtaining light saturation image I between the saturation image J that delusters based on atmospherical scattering model is as follows：
I (x)=J (x) t (x)+A (1-t (x))
T (x)=e-βd(x) (2)
Wherein, θ0、θ1、θ2For linear coefficient, it is random error to take 0.12,0.96 and -0.78, ε respectively, and it is 0 side to take mean value
The Gaussian function that difference is 0.05, A are picture luminance peak, and t is transmission function, and between taking 0.1 to 0.9, β is atmospheric scattering system
Number, takes 1.Finally obtain the saturation image J that delusters：
B, after saturation picture is delustered in step a generation, using Itti significance tests method to input picture first into
The multiple feature channels of row and multiple dimensioned decomposition, establish 9 layers of pyramid, then extract brightness spy respectively to each layer of pyramid
Sign, red feature, green characteristic, blue characteristics, yellow characteristic, direction character form brightness pyramid, chromaticity gold
Word tower and direction character pyramid.
It is poor to make between the pyramidal center scale of different characteristic and edge scale respectively.Brightness figure, face is calculated
Color characteristic figure, direction character figure.
The characteristic pattern of different scale and different characteristic is merged, using normalized function first to different in each feature
The characteristic pattern of scale is normalized, and the synthesis notable figure of a width this feature is formed, then again by the notable of different characteristic
Figure carries out linear, additive and mean value is taken to obtain overall visual notable figure.
C, the mean value m and variance d of notable picture are extracted, calculates m ± 3d as picture conspicuousness segmentation threshold, extraction is notable
The higher pixel region of saliency value in figure carries out binary conversion treatment to original image and obtains the potential region of corncob.
(2) misrecognition region removal：Picture marking area is obtained by step (1), extracts the picture feature of each region
Classify, it is final to identify corncob region；It is as follows：
A, the picture feature in corncob region and background area is extracted, the picture feature includes mean value, standard deviation, smoothly
Degree, third moment, consistency, six grey level histogram features of gray level entropy, contrast, correlation, energy on four direction and same
Matter totally ten six gray level co-occurrence matrixes features.L × Num eigenmatrixes of each picture are obtained, this feature matrix includes L sample
Originally with Num feature, the grey level histogram feature of picture connected region can be expressed asAnd gray scale
Co-occurrence matrix feature
B, training sample is divided into corncob region and background area after the feature that each potential region is calculated in step a
LSSVM disaggregated models are generated, when being tested by LSSVM models, the feature in potential region in test pictures is substituted into LSSVM
It is tested in grader, determines corncob region, delete operation is carried out to background area, obtained corncob and finally identify knot
Fruit.
Claims (10)
1. a kind of corncob automatic identifying method based on significance test, which is characterized in that include the following steps：
(1) the potential region recognition of corncob：
Field picture when S1, acquisition corn are in ear period is as experiment sample；
S2, the depth of view information based on simulated light zone of saturation carry out saturated process of delustering to field picture, obtain the saturation figure that delusters
Piece；
S3, brightness, color characteristic and the side that the saturation each pixel of picture of delustering is calculated using Itti significance test algorithms
Linear superposition is carried out to feature, and by three features, obtains the notable figure after Fusion Features；
S4, the pixel region that saliency value is high in notable figure is extracted using adaptive threshold fuzziness, obtains the potential region of corncob；
(2) misrecognition region removal：
S5, the picture feature for extracting each potential region of corncob；
S6, LSSVM disaggregated models are established, training sample is divided into corncob region and background area；
S7, it determines corncob region, removes background area, obtain the final recognition result of corncob.
2. corncob automatic identifying method according to claim 1, which is characterized in that the S2 the specific steps are：
1) the brightness v and saturation degree s for extracting pixel in the picture of field, establish image depth d with the line between image slices vegetarian refreshments x
Property model：
D (x)=θ0+θ1v(x)+θ2s(x)+ε(x) (1)
2) relationship that light saturation image I is obtained between the saturation image J that delusters based on atmospherical scattering model：
I (x)=J (x) t (x)+A (1-t (x))
T (x)=e-βd(x) (2)
Wherein, θ0、θ1、θ2For linear coefficient, it is random error to take 0.12,0.96 and -0.78, ε respectively, and it is that 0 variance is to take mean value
0.05 Gaussian function, A are picture luminance peak, and t is transmission function, and between taking 0.1 to 0.9, β is atmospheric scattering coefficient,
Take 1；
3) saturation image expression formula of delustering is obtained：
3. corncob automatic identifying method according to claim 1 or 2, which is characterized in that the S3 the specific steps are：
4) 9 layers of gaussian pyramid is established by input picture, wherein the 0th layer is input picture, the 1 to 8th layer is through 5 × 5 respectively
Gaussian filter input picture is filtered and is sampled obtain image, size is respectively the 1/2,1/4,1/ of input picture
8,1/16,1/32,1/64,1/128,1/256；Then brightness, red is extracted respectively to each tomographic image of gaussian pyramid
Color characteristic, green characteristic, blue characteristics, yellow characteristic, direction character form brightness pyramid, chromaticity pyramid
With direction character pyramid；
5) feature takes difference to form characteristic pattern between different scale：Respectively in the pyramidal center scale c of different characteristic and edge ruler
Work is poor between spending s, and brightness figure, color characteristic figure and direction character figure is calculated；Wherein, { 2,3,4 } c ∈, s=c+ δ, δ
∈{3,4}；
6) characteristic pattern of the characteristic pattern of different scale and different characteristic is merged, generates notable figure：Use normalized function
First the characteristic pattern of different scale in each feature is normalized, forms the synthesis notable figure of a width this feature, then will
The notable figure of different characteristic carries out linear, additive and takes mean value, obtains overall visual notable figure.
4. corncob automatic identifying method according to claim 1 or 2, which is characterized in that the S4 the specific steps are：
If the pixel significance distribution of every width notable figure meets Gaussian Profile rule, the mean value m and variance d of each pixel of notable figure are extracted,
M ± 3d is calculated as notable figure conspicuousness segmentation threshold, the higher pixel region of saliency value in notable figure is extracted, to original image
It carries out binary conversion treatment and obtains the potential region of corncob.
5. corncob automatic identifying method according to claim 3, which is characterized in that the S4 the specific steps are：If
The pixel significance distribution of every width notable figure meets Gaussian Profile rule, extracts the mean value m and variance d of each pixel of notable figure, meter
Calculate m ± 3d be used as notable figure conspicuousness segmentation threshold, extraction notable figure in the higher pixel region of saliency value, to original image into
Row binary conversion treatment obtains the potential region of corncob.
6. according to the corncob automatic identifying method described in claim 1,2 or 5, which is characterized in that the picture in the S5 is special
Sign includes grey level histogram featureWith gray level co-occurrence matrixes feature
Include L sample and Num feature；The grey level histogram feature includes mean value, standard deviation, smoothness, third moment, consistent
Six property, gray level entropy features；The gray level co-occurrence matrixes feature includes contrast on four direction, correlation, energy and same
Matter textural characteristics, totally ten six features.
7. corncob automatic identifying method according to claim 3, which is characterized in that the picture feature in the S5 includes
Grey level histogram featureWith gray level co-occurrence matrixes featureInclude L
A sample and Num feature；The grey level histogram feature includes mean value, standard deviation, smoothness, third moment, consistency, gray scale
Six features of entropy；The gray level co-occurrence matrixes feature includes contrast, correlation, energy and homogeney texture on four direction
Feature, totally ten six features.
8. corncob automatic identifying method according to claim 4, which is characterized in that the picture feature in the S5 includes
Grey level histogram featureWith gray level co-occurrence matrixes featureInclude L
A sample and Num feature；The grey level histogram feature includes mean value, standard deviation, smoothness, third moment, consistency, gray scale
Six features of entropy；The gray level co-occurrence matrixes feature includes contrast, correlation, energy and homogeney texture on four direction
Feature, totally ten six features.
9. according to the corncob automatic identifying method described in claim 1,2,5,7 or 8, which is characterized in that the experiment sample
It is the picture that corn is in ear period between 3 years, wherein First Year randomly selects half picture as training sample.
10. corncob automatic identifying method according to claim 6, which is characterized in that the experiment sample is between 3 years
Corn is in the picture of ear period, and wherein First Year randomly selects half picture as training sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810295953.3A CN108537267A (en) | 2018-04-02 | 2018-04-02 | A kind of corncob automatic identifying method based on significance test |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810295953.3A CN108537267A (en) | 2018-04-02 | 2018-04-02 | A kind of corncob automatic identifying method based on significance test |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108537267A true CN108537267A (en) | 2018-09-14 |
Family
ID=63483042
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810295953.3A Pending CN108537267A (en) | 2018-04-02 | 2018-04-02 | A kind of corncob automatic identifying method based on significance test |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108537267A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106290171A (en) * | 2016-07-27 | 2017-01-04 | 内蒙古大学 | Based on SVM and the maize leaf chlorophyll content of regression analysis and phenotypic parameter assay method |
| CN106846334A (en) * | 2017-01-19 | 2017-06-13 | 江南大学 | Field corn plant recognition methods based on Support Vector data description |
| CN106980861A (en) * | 2017-03-31 | 2017-07-25 | 上海电机学院 | A kind of ship method for quickly identifying based on fusion feature |
| CN107341505A (en) * | 2017-06-07 | 2017-11-10 | 同济大学 | A kind of scene classification method based on saliency Yu Object Bank |
-
2018
- 2018-04-02 CN CN201810295953.3A patent/CN108537267A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106290171A (en) * | 2016-07-27 | 2017-01-04 | 内蒙古大学 | Based on SVM and the maize leaf chlorophyll content of regression analysis and phenotypic parameter assay method |
| CN106846334A (en) * | 2017-01-19 | 2017-06-13 | 江南大学 | Field corn plant recognition methods based on Support Vector data description |
| CN106980861A (en) * | 2017-03-31 | 2017-07-25 | 上海电机学院 | A kind of ship method for quickly identifying based on fusion feature |
| CN107341505A (en) * | 2017-06-07 | 2017-11-10 | 同济大学 | A kind of scene classification method based on saliency Yu Object Bank |
Non-Patent Citations (2)
| Title |
|---|
| QINGSONG ZHU等: "A Fast Single Image Haze Removal Algorithm Using Color Attenuation Prior", 《IEEE TRANSACTIONSON IMAGE PROCESSING》 * |
| 胡佳娟: "基于视觉显著模型和支持向量机的织物瑕疵检测方法", 《中国优秀硕士学位论文全文数据库工程科技I辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bhargava et al. | Fruits and vegetables quality evaluation using computer vision: A review | |
| Francis et al. | Identification of leaf diseases in pepper plants using soft computing techniques | |
| Li et al. | Automatic detection of common surface defects on oranges using combined lighting transform and image ratio methods | |
| CN105718945B (en) | Apple picking robot night image recognition method based on watershed and neural network | |
| CN107977671A (en) | A kind of tongue picture sorting technique based on multitask convolutional neural networks | |
| CN109961024A (en) | Wheat weeds in field detection method based on deep learning | |
| CN106845497B (en) | Corn early-stage image drought identification method based on multi-feature fusion | |
| CN106295661A (en) | The plant species identification method of leaf image multiple features fusion and device | |
| CN103065149A (en) | Netted melon fruit phenotype extraction and quantization method | |
| CN109034269A (en) | A kind of bollworm female male imago method of discrimination based on computer vision technique | |
| CN104680524A (en) | Disease diagnosis method for leaf vegetables | |
| Rani et al. | Pest identification in leaf images using SVM classifier | |
| Jadhav et al. | Soybean leaf disease detection and severity measurement using multiclass SVM and KNN classifier | |
| CN103593652B (en) | A kind of cucumber disease recognition methods processed based on cucumber leaves symptomatic picture | |
| Ji et al. | In-field automatic detection of maize tassels using computer vision | |
| Pandey et al. | Non-destructive quality grading of mango (Mangifera Indica L) based on CIELab colour model and size | |
| Liu et al. | Deep learning based research on quality classification of shiitake mushrooms | |
| Singh et al. | Automatic detection of rust disease of Lentil by machine learning system using microscopic images. | |
| Sivasakthi et al. | Plant leaf disease identification using image processing and svm, ann classifier methods | |
| Sara et al. | VegNet: An organized dataset of cauliflower disease for a sustainable agro-based automation system | |
| CN112528726B (en) | Cotton aphid pest monitoring method and system based on spectral imaging and deep learning | |
| Li et al. | Disease recognition of maize leaf based on KNN and feature extraction | |
| Xu | Image processing technology in agriculture | |
| Lubis et al. | Classification of tomato leaf disease and combination extraction features using K-NN algorithm | |
| CN115862003A (en) | Lightweight YOLOv 5-based in-vivo apple target detection and classification method |
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: 20180914 |
|
| RJ01 | Rejection of invention patent application after publication |