CN107239514A - A kind of plants identification method and system based on convolutional neural networks - Google Patents
A kind of plants identification method and system based on convolutional neural networks Download PDFInfo
- Publication number
- CN107239514A CN107239514A CN201710359176.XA CN201710359176A CN107239514A CN 107239514 A CN107239514 A CN 107239514A CN 201710359176 A CN201710359176 A CN 201710359176A CN 107239514 A CN107239514 A CN 107239514A
- Authority
- CN
- China
- Prior art keywords
- plant
- image
- neural networks
- convolutional neural
- recognized
- 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
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 58
- 238000000034 method Methods 0.000 title claims abstract description 26
- 241000196324 Embryophyta Species 0.000 claims abstract description 132
- 238000000605 extraction Methods 0.000 claims abstract description 15
- 241001529246 Platymiscium Species 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 27
- 239000000284 extract Substances 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 230000009467 reduction Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 abstract description 5
- 238000013507 mapping Methods 0.000 description 10
- 210000002569 neuron Anatomy 0.000 description 10
- 238000004422 calculation algorithm Methods 0.000 description 7
- 238000012549 training Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 241000723353 Chrysanthemum Species 0.000 description 4
- 235000007516 Chrysanthemum Nutrition 0.000 description 4
- 238000009826 distribution Methods 0.000 description 4
- 230000007935 neutral effect Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 241000220317 Rosa Species 0.000 description 2
- 210000004027 cell Anatomy 0.000 description 2
- 210000005036 nerve Anatomy 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 244000170916 Paeonia officinalis Species 0.000 description 1
- 235000006484 Paeonia officinalis Nutrition 0.000 description 1
- 235000008331 Pinus X rigitaeda Nutrition 0.000 description 1
- 235000011613 Pinus brutia Nutrition 0.000 description 1
- 241000018646 Pinus brutia Species 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 235000011389 fruit/vegetable juice Nutrition 0.000 description 1
- 230000008571 general function Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Library & Information Science (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of plants identification method and system based on convolutional neural networks, wherein method comprises the following steps：The plant image of collection, plant similar in plant image is classified and marked, obtain plant database；Plant image in plant database is inputted to convolutional neural networks, convolutional neural networks are trained and obtain characteristic matching model；Images to be recognized is received, using the image feature value of characteristic matching model extraction images to be recognized, and image feature value and the similarity of plant in plant database is calculated, the classification of images to be recognized institute platymiscium is judged according to similarity.The present invention completes the adaptivity of plants identification by convolutional neural networks, reduces design difficulty and amount of calculation, improves the efficiency of identification plant.
Description
Technical field
The present invention relates to plants identification technical field, and in particular to a kind of plants identification method based on convolutional neural networks
And system.
Background technology
It is usually feature (such as SIFT, HOG, LBP for extracting default plant image in advance in classical pattern-recognition
Feature).Extract after feature, feature is encoded, such as commonly use BoW, FisherVector etc..Then feature is put into one
Grader, such as SVM, carry out 2 and classify, train optimal classification surface, the feature of certain class plant can most be represented by finding, remove pair
The unrelated and autocorrelative feature of classification.However, the extraction of these features too relies on the experience and subjective consciousness of people, extract
The different of feature influence very big to classification performance, or even the order for the feature extracted can also influence last plant classification performance.
Meanwhile, the quality of image preprocessing also influences whether the feature extracted.
The content of the invention
The step of purpose of the present invention is simplified plants identification, reduces the amount of calculation of plants identification process and improves plant knowledge
Other precision.
To achieve these goals, the invention provides a kind of plants identification method based on convolutional neural networks, including
Following steps：
The plant image of collection, plant similar in plant image is classified and marked, obtain plant database；
Plant image in plant database is inputted to convolutional neural networks, convolutional neural networks are trained and obtain spy
Levy Matching Model；
Images to be recognized is received, using the image feature value of characteristic matching model extraction images to be recognized, and image is calculated
Characteristic value and the similarity of plant in plant database, the classification of images to be recognized institute platymiscium is judged according to similarity.
Further, in the above-mentioned plants identification method based on convolutional neural networks, convolutional neural networks include defeated successively
Enter layer, convolution operation layer and output layer, wherein,
Input layer, for receiving plant image；
Convolution operation layer includes convolutional layer and pond layer, wherein,
Convolutional layer, for extracting image characteristic matrix according to convolution kernel；
Pond layer, the image of this plant image local feature can be most represented for extracting in each image characteristic matrix
Characteristic value.
Further, in the above-mentioned plants identification method based on convolutional neural networks, the reception images to be recognized is utilized
The image feature value of characteristic matching model extraction images to be recognized, and calculate image feature value and the phase of plant in plant database
Like spending, the classification for judging images to be recognized institute platymiscium according to similarity includes
N number of image characteristic matrix is extracted using convolution kernel in characteristic matching model, and is extracted in each image characteristic matrix
Image feature value；
Eigenvalue matrix is worth to according to all characteristics of image；
Eigenvalue matrix and the similarity of plant in plant database are calculated using default function.
Further, it is described by plant in plant database in the above-mentioned plants identification method based on convolutional neural networks
Image is inputted also to be included to before convolutional neural networks
Plant image is pre-processed, pretreatment at least includes normalization, lightness adjustment or noise reduction.
Further, in the above-mentioned plants identification method based on convolutional neural networks, image feature value at least includes color
Value, profile value, marginal value, gray value.
Present invention also offers a kind of Plant identification based on convolutional neural networks, including
Receiving module, plant similar in plant image for the plant image of collection, classified and marked, obtained
Plant database；
Characteristic extracting module, for plant image in plant database to be inputted to convolutional neural networks, to convolutional Neural
Network, which is trained, obtains characteristic matching model；
Identification module, for receiving images to be recognized, utilizes the characteristics of image of characteristic matching model extraction images to be recognized
Value, and image feature value and the similarity of plant in plant database are calculated, planted according to belonging to similarity judges images to be recognized
The classification of thing.
Further, in the above-mentioned Plant identification based on convolutional neural networks, the identification module includes
Convolution unit, for extracting N number of image characteristic matrix using convolution kernel in characteristic matching model, and extracts each figure
As image feature value in eigenmatrix；
Pond unit, for being worth to eigenvalue matrix according to all characteristics of image；
Computing unit, for calculating eigenvalue matrix and the similarity of plant in plant database using default function.
Further, in the above-mentioned Plant identification based on convolutional neural networks, the characteristic extracting module includes
Subelement is pre-processed, for being pre-processed to plant image, pretreatment at least includes normalization, lightness adjustment
Or noise reduction.
In the above-mentioned technical solutions, it is of the invention with must first find the profile or edge of leaf in traditional recognition methods
Etc. entirely different, without defining any edge or profile；The present invention avoids explicit feature using convolutional neural networks and taken
Sample, is implicitly learnt from training data.This causes convolutional neural networks to be substantially different from other based on neutral net
Feature extraction functions are integrated into multilayer perceptron by grader by structural rearrangement and reduction weights.It can directly handle ash
Picture is spent, can be directly used for handling the classification based on image.
Brief description of the drawings
, below will be to institute in embodiment in order to illustrate more clearly of the embodiment of the present application or technical scheme of the prior art
The accompanying drawing needed to use is briefly described, it should be apparent that, drawings in the following description are only one described in the present invention
A little embodiments, for those of ordinary skill in the art, can also obtain other accompanying drawings according to these accompanying drawings.
Fig. 1 illustrates for the flow of plants identification method one embodiment of the present invention based on convolutional neural networks
Figure；
Fig. 2 is in plants identification method one embodiment of the present invention based on convolutional neural networks
S103 schematic flow sheet；
Fig. 3 is in plants identification method one embodiment of the present invention based on convolutional neural networks
Feature extraction and calculation schematic diagram in S103；
Fig. 4 illustrates for the structure of block diagram of botanical system one embodiment of the present invention based on convolutional neural networks
Figure.
Embodiment
In order that those skilled in the art more fully understands technical scheme, below in conjunction with accompanying drawing to this hair
It is bright to be further detailed.
As shown in figure 1, the invention provides a kind of plants identification method based on convolutional neural networks, including following step
Suddenly：
S101, the plant image of collection, plant similar in plant image is classified and marked, obtain plant data
Storehouse；When it is implemented, the image of collection is optionally shot using the equipment including camera, such as mobile phone, ipad.
The step purpose is the image for collecting enough plant, and each image is marked (indicated each
Existing in plant image in which class plant, such as image has rose, or has pine tree etc.), so as to form each floristic
(such as this class of rose has 200 photos to image set, and peony has 100 pictures etc., and a plant is correspond to per pictures
Species).
S102, the plant image after classification and mark inputted to convolutional neural networks, convolutional neural networks are instructed
Get Image Feature Matching model；
Heretofore described convolutional neural networks include input layer, convolution operation layer and output layer successively, wherein,
Input layer, for receiving plant image；
Convolution operation layer includes convolutional layer and pond layer, wherein,
Convolutional layer, for extracting image characteristic matrix according to convolution kernel；
Pond layer, the image of this plant image local feature can be most represented for extracting in each image characteristic matrix
Characteristic value.
Training process：By the plant image collected in process S101 by batch input (the CNN networks into CNN networks
Need into row stochastic initialization), the gap (being referred to as Loss) between the batch and target is calculated, then according to this
Loss carries out seeking partial differential for the convolution kernel (i.e. parameter) in network, and calculating each convolution kernel (can specifically use general
Function such as Sigmoid, ReLU etc. are calculated) gap between target, and related parameter (reference learning ratio) is entered
Row updates, so as to update whole CNN neutral nets；Change image to input into CNN networks, according to batch set in advance, instruction
Practice number of times etc. to repeat, the relevant parameter of whole neutral net is updated, a complete nerve is to the last obtained
Network (including related path and parameter), that is, obtain the CNN trained a neural network model.
S103, reception images to be recognized, using the image feature value of characteristic matching model extraction images to be recognized, and are calculated
Image feature value and the similarity of plant in plant database, the classification of images to be recognized institute platymiscium is judged according to similarity.
Traditional mode classification is almost all based on statistical nature, and this means that must extract certain before being differentiated
A little features.However, explicit feature extraction is not easy to, it is also and not always reliable in some application problems.Profit of the invention
Had the following advantages with the more general neutral net of convolutional neural networks in terms of image procossing：A) the topology knot of input picture and network
Structure can coincide well；B) feature extraction and pattern classification are carried out simultaneously, and are produced simultaneously in training；C) weight is shared can be with
The training parameter of network is reduced, neural network structure is become simpler, adaptability is stronger.
Further, S103 includes when being embodied as shown in Figure 2
S1031, using convolution kernel in characteristic matching model N number of image characteristic matrix is extracted, and extract each characteristics of image
Image feature value in matrix；Further, image feature value includes color value, profile value, marginal value, gray value etc..
S1032, according to all characteristics of image it is worth to eigenvalue matrix；
S1033, the similarity using plant in default function calculating eigenvalue matrix and plant database.
Now understood with schematic diagram as shown in Figure 3, the change change poles that the present invention passes through image characteristic matrix to eigenvalue matrix
The big amount of calculation for reducing Similarity Measure process, improves calculating speed.
When being trained in the present invention using plant image to convolutional neural networks, plant image can be divided first
Class, the image of each plant is split, and is divided into training with image and verification image.Verification image is input to nerve
In network model, corresponding feature is extracted, then be compared with the algorithm path in convolutional neural networks with parameter sets, root
Go out the similarity of this feature and the botanical name according to box counting algorithm, so that the accuracy of testing model；If model is accurate
Degree is low, and convolutional neural networks can be trained again, deeper to go the characteristic relation for extracting these images (more
It is algorithm path and parameter), form new model (including the set of algorithm path+parameter)；After the model updated,
In use, directly carrying out aspect ratio pair to the data newly entered such as image according to the algorithm in model and parameter, correlation is obtained
Similarity value.
Illustrate that convolutional neural networks workflow is as follows in the present invention exemplified by receiving image below, input layer is by 32 × 32
Individual sensing node composition, receives original image.Then, calculation process between convolution and sub-sample alternately, as described below：
First hidden layer carries out convolution, and it is made up of 8 Feature Mappings, and each Feature Mapping is by 28 × 28 neural tuples
Into each neuron specifies the acceptance region of one 5 × 5, and this 28 × 28 neurons share 5 × 5 weighting parameters, i.e. convolution
Core；
Second hidden layer realizes sub-sample and local average, and it is equally made up of 8 Feature Mappings, but each of which feature is reflected
Penetrate and be made up of 14 × 14 neurons.Each neuron has the acceptance region of one 2 × 2, and one can train coefficient, and one can instruct
Practice biasing and a sigmoid activation primitive.The operating point of coefficient and biasing control neuron can be trained；
3rd hidden layer carries out second of convolution, and it is made up of 20 Feature Mappings, and each Feature Mapping is by 10 × 10
Neuron is constituted.Each neuron in the hidden layer may have the cynapse being connected with the several Feature Mappings of next hidden layer
Connection, it to first similar mode of convolutional layer to operate.
4th hidden layer carries out second of sub-sample and local average juice is calculated.It is made up of 20 Feature Mappings, but often
Individual Feature Mapping is made up of 5 × 5 neurons, and it to first time similar mode of sampling to operate.
5th hidden layer realizes the final stage of convolution, and it is made up of 120 neurons, and each neuron specifies one
5 × 5 acceptance region.
It is finally that a full connection obtains output vector.
Continuous alternating of the successive computation layer between convolution and sampling, we obtain the effect of one " double pointed tower ",
It is exactly that, as spatial resolution declines, the quantity of Feature Mapping increases compared with corresponding preceding layer in each convolution or sampling layer
Plus.The thought that sub-sample is carried out after convolution is followed by " complicated " by " simple " cell in animal vision system
The inspiration of the idea of cell and produce.
Further, the S101 also includes
Plant image is pre-processed, pretreatment at least includes normalization, lightness adjustment or noise reduction.
Due to a variety of causes in image, during shooting or transmitting compression, regular noise has been brought into, can be with
Noise reduction process is carried out to image by technological means, it is therefore an objective to increase the description for characteristic in image, highlight feature, make
Feature is more distinct.This means are advance settings, similar with the part function of convolution, i.e., believe the image of plant with classification
Breath is input in neural network model, image is handled (such as the processing of image by the algorithm in model, handling process
Include gray scale, cutting, Feature Mapping etc.).The processing is that training in same category image is handled with image, is found
Such image is than more consistent feature (this feature, which differs, to be set to common to all images), according to the result of processing, generation correspondence
Algorithm path and parameter sets.
As shown in figure 4, the invention provides a kind of Plant identification based on convolutional neural networks, including receiving module
10th, characteristic extracting module 20, identification module 30.Wherein,
Receiving module 10, plant similar in plant image for the plant image of collection, classified and marked, obtained
To plant database；
Characteristic extracting module 20, for plant image in plant database to be inputted to convolutional neural networks, to convolution god
It is trained through network and obtains characteristic matching model；
Identification module 30, it is special using the image of characteristic matching model extraction images to be recognized for receiving images to be recognized
Value indicative, and image feature value and the similarity of plant in plant database are calculated, according to belonging to similarity judges images to be recognized
The classification of plant.
The different photos of 100 chrysanthemums are gathered, photo is divided into the training set A and 10 image verification set B of 90；
90 images are separately input in convolutional neural networks model first, it is assumed that can be done in convolutional neural networks model
Following processing：
A) each image is cut into 16*16=256 parts of images
B) gray scale of each image is extracted；
C) color of each image is extracted；
D) edge feature of each image etc. is extracted
E) data preparation of essential characteristic is completed more than
F) in first time trains, the gray value for the small image that model is cut out to each image is compared, system
The intensity profile for counting out each small image compares；
G) in second is trained, the color-values to each small image are compared, and counting color-values must be distributed；
H) in third time is trained, counting edge feature is worth distribution；
I) repeatedly train, by carrying out the segmentation of more details, feature extraction, so as to form the plant to image
Distinctive processing path, parameter sets (i.e. feature distribution)；
Secondly, whether the characteristics of image identification model that 10 images can be used for generating forms the distinctive path of the plant
Set and parameter sets；
Finally, whether be chrysanthemum, then the image is input into model, according to Models Sets if judging the flower in a unknown images
The route characteristic and parameter sets of each class plant in conjunction, calculate the plant and similarity (phase in model per class plant
It is the statistical result of the plant different dimensions feature weight and statistical distribution in fact like degree, such as the gray value of chrysanthemum is 100, system
Distributed areas are counted for (80~120), normal distribution, 0.01) gray feature weight in the identification model of whole chrysanthemum is.
Some one exemplary embodiments of the present invention are only described by way of explanation above, undoubtedly, for ability
The those of ordinary skill in domain, without departing from the spirit and scope of the present invention, can be with a variety of modes to institute
The embodiment of description is modified.Therefore, above-mentioned accompanying drawing and description are inherently illustrative, should not be construed as to the present invention
The limitation of claims.
Claims (8)
1. a kind of plants identification method based on convolutional neural networks, it is characterised in that comprise the following steps：
The plant image of collection, plant similar in plant image is classified and marked, obtain plant database；
Plant image in plant database is inputted to convolutional neural networks, convolutional neural networks are trained and obtain feature
With model；
Images to be recognized is received, using the image feature value of characteristic matching model extraction images to be recognized, and characteristics of image is calculated
Value and the similarity of plant in plant database, the classification of images to be recognized institute platymiscium is judged according to similarity.
2. the plants identification method according to claim 1 based on convolutional neural networks, it is characterised in that the convolution god
Include input layer, convolution operation layer and output layer successively through network, wherein,
Input layer, for receiving plant image；
Convolution operation layer includes convolutional layer and pond layer, wherein,
Convolutional layer, for extracting image characteristic matrix according to convolution kernel；
Pond layer, the characteristics of image of this plant image local feature can be most represented for extracting in each image characteristic matrix
Value.
3. the plants identification method according to claim 1 based on convolutional neural networks, it is characterised in that the reception is treated
Image is recognized, using the image feature value of characteristic matching model extraction images to be recognized, and image feature value and plant number is calculated
According to the similarity of plant in storehouse, the classification for judging images to be recognized institute platymiscium according to similarity includes
N number of image characteristic matrix is extracted using convolution kernel in characteristic matching model, and extracts image in each image characteristic matrix
Characteristic value；
Eigenvalue matrix is worth to according to all characteristics of image；
Eigenvalue matrix and the similarity of plant in plant database are calculated using default function.
4. the plants identification method according to claim 1 based on convolutional neural networks, it is characterised in that described by plant
Plant image is inputted in database also includes to before convolutional neural networks
Plant image is pre-processed, pretreatment at least includes normalization, lightness adjustment or noise reduction.
5. the plants identification method according to claim 1 based on convolutional neural networks, it is characterised in that image feature value
At least include color value, profile value, marginal value, gray value.
6. a kind of Plant identification based on convolutional neural networks, it is characterised in that including
Receiving module, plant similar in plant image for the plant image of collection, classified and marked, obtain plant
Database；
Characteristic extracting module, for plant image in plant database to be inputted to convolutional neural networks, to convolutional neural networks
It is trained and obtains characteristic matching model；
Identification module, for receiving images to be recognized, using the image feature value of characteristic matching model extraction images to be recognized, and
Image feature value and the similarity of plant in plant database are calculated, point of images to be recognized institute platymiscium is judged according to similarity
Class.
7. the Plant identification according to claim 6 based on convolutional neural networks, it is characterised in that the identification mould
Block includes
Convolution unit, for extracting N number of image characteristic matrix using convolution kernel in characteristic matching model, and it is special to extract each image
Levy image feature value in matrix；
Pond unit, for being worth to eigenvalue matrix according to all characteristics of image；
Computing unit, for calculating eigenvalue matrix and the similarity of plant in plant database using default function.
8. the Plant identification according to claim 6 based on convolutional neural networks, it is characterised in that the feature is carried
Modulus block includes
Subelement is pre-processed, for being pre-processed to plant image, pretreatment at least includes normalization, lightness adjustment or dropped
Make an uproar.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710359176.XA CN107239514A (en) | 2017-05-19 | 2017-05-19 | A kind of plants identification method and system based on convolutional neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710359176.XA CN107239514A (en) | 2017-05-19 | 2017-05-19 | A kind of plants identification method and system based on convolutional neural networks |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107239514A true CN107239514A (en) | 2017-10-10 |
Family
ID=59984520
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710359176.XA Pending CN107239514A (en) | 2017-05-19 | 2017-05-19 | A kind of plants identification method and system based on convolutional neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107239514A (en) |
Cited By (22)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107957859A (en) * | 2017-12-12 | 2018-04-24 | 歌尔科技有限公司 | Image presentation method and virtual reality device for virtual reality child teaching |
| CN108171243A (en) * | 2017-12-18 | 2018-06-15 | 广州七乐康药业连锁有限公司 | A kind of medical image information recognition methods and system based on deep neural network |
| CN108256568A (en) * | 2018-01-12 | 2018-07-06 | 宁夏智启连山科技有限公司 | A kind of plant species identification method and device |
| CN109165611A (en) * | 2018-08-31 | 2019-01-08 | 广州市派客朴食信息科技有限责任公司 | A kind of dishes identification intelligent settlement method based on machine vision and neural network |
| CN109446985A (en) * | 2018-10-28 | 2019-03-08 | 贵州师范学院 | Multi-angle plants identification method based on vector neural network |
| CN109522797A (en) * | 2018-10-16 | 2019-03-26 | 华南农业大学 | Rice seedling and Weeds at seedling recognition methods and system based on convolutional neural networks |
| CN109583301A (en) * | 2018-10-29 | 2019-04-05 | 广东奥博信息产业股份有限公司 | A kind of optimal external planting conditions prediction technique of plant growing process and device |
| CN109685050A (en) * | 2018-11-12 | 2019-04-26 | 平安科技（深圳）有限公司 | Character recognition method, device, equipment and storage medium |
| CN110726898A (en) * | 2018-07-16 | 2020-01-24 | 北京映翰通网络技术股份有限公司 | Power distribution network fault type identification method |
| CN110895036A (en) * | 2019-11-30 | 2020-03-20 | 宁波大龙农业科技有限公司 | Automatic temperature and humidity adjusting system of seedling raising greenhouse |
| CN111161237A (en) * | 2019-12-27 | 2020-05-15 | 中山德著智能科技有限公司 | Fruit and vegetable surface quality detection method, storage medium and sorting device thereof |
| CN111345735A (en) * | 2018-12-24 | 2020-06-30 | 江苏美的清洁电器股份有限公司 | Map construction method and device of sweeper |
| CN111652165A (en) * | 2020-06-08 | 2020-09-11 | 北京世纪好未来教育科技有限公司 | Mouth shape evaluating method, mouth shape evaluating equipment and computer storage medium |
| WO2020181817A1 (en) * | 2019-03-12 | 2020-09-17 | 平安科技（深圳）有限公司 | Plant image recognition method and apparatus based on neural network model |
| CN111898680A (en) * | 2020-07-31 | 2020-11-06 | 陈艳 | Biological identification method based on material inspection multi-view morphological image and deep learning |
| CN111950409A (en) * | 2020-07-31 | 2020-11-17 | 江苏大学 | Intelligent identification method and system for road marking line |
| CN112434579A (en) * | 2020-11-13 | 2021-03-02 | 深圳园林股份有限公司 | Method and device for extracting fallen leaf accommodating capacity of ground cover plant based on neural network |
| CN112699901A (en) * | 2021-01-07 | 2021-04-23 | 镇江瑞奇信息技术有限公司 | Plant species identification system based on Internet of things |
| CN112861900A (en) * | 2019-11-28 | 2021-05-28 | 财团法人资讯工业策进会 | Image system and detection method |
| CN112990315A (en) * | 2021-03-17 | 2021-06-18 | 北京大学 | 3D shape image classification method of equal-variation 3D convolution network based on partial differential operator |
| CN114451180A (en) * | 2022-02-14 | 2022-05-10 | 四川上太科技有限公司 | Photovoltaic-based greenhouse system |
| US11853368B2 (en) | 2019-07-10 | 2023-12-26 | Hangzhou Glority Software Limited | Method and system for identifying and displaying an object |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106022285A (en) * | 2016-05-30 | 2016-10-12 | 北京智芯原动科技有限公司 | Vehicle type identification method and vehicle type identification device based on convolutional neural network |
| US20160307070A1 (en) * | 2013-12-27 | 2016-10-20 | Huawei Technologies Co., Ltd. | Method for Recognizing Target Object in Image, and Apparatus |
| CN106529446A (en) * | 2016-10-27 | 2017-03-22 | 桂林电子科技大学 | Vehicle type identification method and system based on multi-block deep convolutional neural network |
| CN106682704A (en) * | 2017-01-20 | 2017-05-17 | 中国科学院合肥物质科学研究院 | Method of disease image identification based on hybrid convolutional neural network fused with context information |
-
2017
- 2017-05-19 CN CN201710359176.XA patent/CN107239514A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160307070A1 (en) * | 2013-12-27 | 2016-10-20 | Huawei Technologies Co., Ltd. | Method for Recognizing Target Object in Image, and Apparatus |
| CN106022285A (en) * | 2016-05-30 | 2016-10-12 | 北京智芯原动科技有限公司 | Vehicle type identification method and vehicle type identification device based on convolutional neural network |
| CN106529446A (en) * | 2016-10-27 | 2017-03-22 | 桂林电子科技大学 | Vehicle type identification method and system based on multi-block deep convolutional neural network |
| CN106682704A (en) * | 2017-01-20 | 2017-05-17 | 中国科学院合肥物质科学研究院 | Method of disease image identification based on hybrid convolutional neural network fused with context information |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107957859A (en) * | 2017-12-12 | 2018-04-24 | 歌尔科技有限公司 | Image presentation method and virtual reality device for virtual reality child teaching |
| CN108171243A (en) * | 2017-12-18 | 2018-06-15 | 广州七乐康药业连锁有限公司 | A kind of medical image information recognition methods and system based on deep neural network |
| CN108256568A (en) * | 2018-01-12 | 2018-07-06 | 宁夏智启连山科技有限公司 | A kind of plant species identification method and device |
| CN110726898A (en) * | 2018-07-16 | 2020-01-24 | 北京映翰通网络技术股份有限公司 | Power distribution network fault type identification method |
| CN110726898B (en) * | 2018-07-16 | 2022-02-22 | 北京映翰通网络技术股份有限公司 | Power distribution network fault type identification method |
| CN109165611A (en) * | 2018-08-31 | 2019-01-08 | 广州市派客朴食信息科技有限责任公司 | A kind of dishes identification intelligent settlement method based on machine vision and neural network |
| CN109165611B (en) * | 2018-08-31 | 2020-03-24 | 广州市派客朴食信息科技有限责任公司 | Intelligent dish identification settlement method based on machine vision and neural network |
| CN109522797A (en) * | 2018-10-16 | 2019-03-26 | 华南农业大学 | Rice seedling and Weeds at seedling recognition methods and system based on convolutional neural networks |
| CN109446985A (en) * | 2018-10-28 | 2019-03-08 | 贵州师范学院 | Multi-angle plants identification method based on vector neural network |
| CN109446985B (en) * | 2018-10-28 | 2021-06-04 | 贵州师范学院 | Multi-angle plant identification method based on vector neural network |
| CN109583301A (en) * | 2018-10-29 | 2019-04-05 | 广东奥博信息产业股份有限公司 | A kind of optimal external planting conditions prediction technique of plant growing process and device |
| CN109685050A (en) * | 2018-11-12 | 2019-04-26 | 平安科技（深圳）有限公司 | Character recognition method, device, equipment and storage medium |
| CN111345735A (en) * | 2018-12-24 | 2020-06-30 | 江苏美的清洁电器股份有限公司 | Map construction method and device of sweeper |
| WO2020181817A1 (en) * | 2019-03-12 | 2020-09-17 | 平安科技（深圳）有限公司 | Plant image recognition method and apparatus based on neural network model |
| US11853368B2 (en) | 2019-07-10 | 2023-12-26 | Hangzhou Glority Software Limited | Method and system for identifying and displaying an object |
| CN112861900B (en) * | 2019-11-28 | 2023-08-01 | 财团法人资讯工业策进会 | Image system and detection method |
| CN112861900A (en) * | 2019-11-28 | 2021-05-28 | 财团法人资讯工业策进会 | Image system and detection method |
| CN110895036A (en) * | 2019-11-30 | 2020-03-20 | 宁波大龙农业科技有限公司 | Automatic temperature and humidity adjusting system of seedling raising greenhouse |
| CN111161237A (en) * | 2019-12-27 | 2020-05-15 | 中山德著智能科技有限公司 | Fruit and vegetable surface quality detection method, storage medium and sorting device thereof |
| CN111652165A (en) * | 2020-06-08 | 2020-09-11 | 北京世纪好未来教育科技有限公司 | Mouth shape evaluating method, mouth shape evaluating equipment and computer storage medium |
| CN111898680A (en) * | 2020-07-31 | 2020-11-06 | 陈艳 | Biological identification method based on material inspection multi-view morphological image and deep learning |
| CN111950409A (en) * | 2020-07-31 | 2020-11-17 | 江苏大学 | Intelligent identification method and system for road marking line |
| CN111950409B (en) * | 2020-07-31 | 2024-05-14 | 江苏大学 | Intelligent identification method and system for road marking line |
| CN112434579A (en) * | 2020-11-13 | 2021-03-02 | 深圳园林股份有限公司 | Method and device for extracting fallen leaf accommodating capacity of ground cover plant based on neural network |
| CN112699901A (en) * | 2021-01-07 | 2021-04-23 | 镇江瑞奇信息技术有限公司 | Plant species identification system based on Internet of things |
| CN112990315B (en) * | 2021-03-17 | 2023-10-20 | 北京大学 | 3D shape image classification method of constant-variation 3D convolution network based on partial differential operator |
| CN112990315A (en) * | 2021-03-17 | 2021-06-18 | 北京大学 | 3D shape image classification method of equal-variation 3D convolution network based on partial differential operator |
| CN114451180A (en) * | 2022-02-14 | 2022-05-10 | 四川上太科技有限公司 | Photovoltaic-based greenhouse system |
| CN114451180B (en) * | 2022-02-14 | 2023-12-12 | 四川上太科技有限公司 | Greenhouse system based on photovoltaic |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107239514A (en) | A kind of plants identification method and system based on convolutional neural networks | |
| CN108717568B (en) | A kind of image characteristics extraction and training method based on Three dimensional convolution neural network | |
| Swasono et al. | Classification of tobacco leaf pests using VGG16 transfer learning | |
| CN109308697B (en) | Leaf disease identification method based on machine learning algorithm | |
| Kumar et al. | Resnet-based approach for detection and classification of plant leaf diseases | |
| CN111476713B (en) | Intelligent weather image identification method and system based on multi-depth convolution neural network fusion | |
| CN111274860B (en) | Recognition method for online automatic tobacco grade sorting based on machine vision | |
| CN105809198B (en) | SAR image target recognition method based on depth confidence network | |
| CN112734775B (en) | Image labeling, image semantic segmentation and model training methods and devices | |
| CN109101938B (en) | Multi-label age estimation method based on convolutional neural network | |
| CN112308158A (en) | Multi-source field self-adaptive model and method based on partial feature alignment | |
| CN106446942A (en) | Crop disease identification method based on incremental learning | |
| CN110517311A (en) | Pest and disease monitoring method based on leaf spot lesion area | |
| CN104143079A (en) | Method and system for face attribute recognition | |
| CN106600595A (en) | Human body characteristic dimension automatic measuring method based on artificial intelligence algorithm | |
| CN110490227A (en) | A kind of few sample image classification method based on Feature Conversion | |
| CN110479636A (en) | Method and device based on neural network automatic sorting tobacco leaf | |
| CN109886295A (en) | A kind of butterfly recognition methods neural network based and relevant device | |
| CN104063686A (en) | System and method for performing interactive diagnosis on crop leaf segment disease images | |
| CN109101891A (en) | A kind of rice pest detection system and its detection method merging artificial intelligence | |
| CN106874825A (en) | The training method of Face datection, detection method and device | |
| CN110096976A (en) | Human behavior micro-Doppler classification method based on sparse migration network | |
| Bhimavarapu et al. | Analysis and characterization of plant diseases using transfer learning | |
| CN111126155B (en) | Pedestrian re-identification method for generating countermeasure network based on semantic constraint | |
| Borman et al. | Classification of Medicinal Wild Plants Using Radial Basis Function Neural Network with Least Mean Square |
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: 20171010 |