CN106372648A - Multi-feature-fusion-convolutional-neural-network-based plankton image classification method - Google Patents
Multi-feature-fusion-convolutional-neural-network-based plankton image classification method Download PDFInfo
- Publication number
- CN106372648A CN106372648A CN201610912684.1A CN201610912684A CN106372648A CN 106372648 A CN106372648 A CN 106372648A CN 201610912684 A CN201610912684 A CN 201610912684A CN 106372648 A CN106372648 A CN 106372648A
- Authority
- CN
- China
- Prior art keywords
- image
- plankton
- convolutional neural
- multiple features
- feature
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
- G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
-
- 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
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- 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]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Probability & Statistics with Applications (AREA)
- Artificial Intelligence (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention provides a multi-feature-fusion-convolutional-neural-network-based plankton image classification method. The method comprises: lots of clear plankton images are collected and a large-scale multi-type plankton image data set is constructed; a global feature and a local feature are extracted by using an image conversion and edge extraction algorithm; an original feature image, a global feature image, and a local feature image are inputted into a depth-learning multi-feature-fusion convolutional neural network to carry out training, thereby obtaining a multi-feature-fusion convolutional neural network model; and then the plankton images are inputted into the multi-feature-fusion convolutional neural network model and classification is carried out based on a finally outputted probability score. According to the invention, the angle of biological morphology, the computer vision method, and the depth learning technology are combined; and thus the classification accuracy for plankton images, especially large-scale multi-type plankton images is high.
Description
Technical field
The present invention relates to biological morphology analysis, computer vision and deep learning technology field are and in particular to a kind of base
Plankton image classification method in multiple features fusion convolutional neural networks.
Background technology
Due to importance in ecosystem for the plankton, the process of plankton image and analysis become increasingly to weigh
Will.However, because planktonic number of species is huge, and the plankton of various species there is also at aspects such as morphological characteristics
Very big difference.For plankton image, same kind of plankton, its profile is not necessarily identical, may
There is very big difference, and for different types of plankton, the feature such as its profile is likely to there is high similarity.This
Similarity between diversity and class in class, brings huge difficult problem to plankton image classification.Traditional image classification method is main
The method being combined using the design of feature extraction and classifying device, but common feature extracting method is not particularly suited for swimming of complexity
Biometric image, and special feature extracting method needs to take a substantial amount of time to carry out research design with energy, and advise for big
The classification of the multi-class plankton image of mould can not obtain good result.
Content of the invention
The application passes through to provide a kind of plankton image classification method based on multiple features fusion convolutional neural networks, with
Solve the difficult technical problem of multi-class on a large scale plankton image classification in prior art.
For solving above-mentioned technical problem, the application employs the following technical solutions and is achieved:
A kind of plankton image classification method based on multiple features fusion convolutional neural networks, comprises the steps:
S1: collection clearly plankton image, builds plankton image data set multi-class on a large scale, wherein,
Plankton image in this data set is as primitive character image;
S2: process primitive character image, extract planktonic global characteristics, obtain global characteristics image, concrete process
Step is:
S21: using image segmentation scharr operator, primitive character image is changed, the image after conversion includes the overall situation
Feature and local feature;
S22: remove the local feature in transition diagram picture using bilateral filtering method;
S23: enhancing contrast ratio is projecting the global characteristics in transition diagram picture;
S3: by the canny edge detection algorithm of computer vision, primitive character image is processed, extract life of swimming
The Edge texture feature of thing, i.e. planktonic local feature, obtain local feature image;
S4: build the multiple features fusion convolutional neural networks model based on primitive character, global characteristics and local feature, should
Multiple features fusion convolutional neural networks include three separate basic sub-networks, and each basic sub-network is respectively trained original
Characteristic image, global characteristics image and local feature image, wherein, 1 to 5 layers of this multiple features fusion convolutional neural networks is volume
Lamination, 6 to 8 layers is full articulamentum;
S5: whole primitive character images, global characteristics image and the local that step s1, step s2 and step s3 are obtained
Characteristic image is input in this multiple features fusion convolutional neural networks model and is trained, and finally gives the multiple features after optimization and melts
Conjunction convolutional neural networks model:
S51: initial state information is set first, including iterationses, learning rate and initialization mode;
S52: this multiple features fusion convolutional neural networks model is carried out with fl transmission and feeds back with backward, make this multiple features melt
Close convolutional neural networks model to be trained according to the plankton image of input and learn;
S53: output loss function value and accuracy rate;
S54: lift the performance of this multiple features fusion convolutional neural networks model by reducing loss function value;
S54: judge whether to reach the iterationses of setting, if it is, training finishes, the multiple features after being optimized melt
Close convolutional neural networks model, otherwise, continue to redirect execution step s52;
S6: plankton image to be sorted is input in the multiple features fusion convolutional neural networks model after optimization,
According to the probability score of final output, judge the classification corresponding to plankton image.
Further, according to practical situation and demand, described basis sub-network can using alexnet, vggnet or
Any one convolutional neural networks in googlenet.Accurate based on the final classification of multiple features fusion convolutional neural networks model
Really rate can step up according to the difference of selected basis sub-network, and correspondingly, the time cost of model training also can be progressively
Increase.
It is that multiple Feature Mapping figures are directly merged mostly in prior art, in order that three kinds of features preferably merge, fill
Divide and excavate with high-dimensional and level information, as a kind of preferred technical scheme, handed over using full connection in described full articulamentum
Fork mixed method merges the Feature Mapping figure that three sub- network trainings obtain.
Differ greatly from view of between global characteristics image and local feature image, this full connection mixing together side
Method, compared with commonly full articulamentum directly merges, effectively reduces global characteristics image and local feature image co-registration is brought
Error, realizes the abundant fusion of multiple features, lifts plankton image classification accuracy rate.
Compared with prior art, the technical scheme that the application provides, the technique effect having or advantage are: from biomorph
Learn angle to set out with computer vision methods, combine with deep learning technology, realize plankton figure multi-class on a large scale
The classification of picture, drastically increases the accuracy rate of classification.
Brief description
Fig. 1 is method of the present invention flow chart;
Fig. 2 is the multiple features fusion convolutional neural networks illustraton of model of the present invention.
Specific embodiment
The embodiment of the present application is passed through to provide a kind of plankton image classification based on multiple features fusion convolutional neural networks
Method, to solve the technical problem of multi-class on a large scale plankton image classification hardly possible in prior art.
In order to be better understood from technique scheme, below in conjunction with Figure of description and specific embodiment, right
Technique scheme is described in detail.
Embodiment
A kind of plankton image classification method based on multiple features fusion convolutional neural networks, as shown in figure 1, include as
Lower step:
S1: collection clearly plankton image, builds plankton image data set multi-class on a large scale, wherein,
, as primitive character image, the amount of images of collection is about 30,000 to 90,000 for plankton image in this data set
, planktonic classification is about in 30 to 50 classes;
S2: process primitive character image, extract planktonic global characteristics, obtain global characteristics image, concrete process
Step is:
S21: using image segmentation scharr operator, primitive character image is changed, the image after conversion includes the overall situation
Feature and local feature;
S22: remove the local feature in transition diagram picture using bilateral filtering method;
S23: enhancing contrast ratio is projecting the global characteristics in transition diagram picture;
S3: by the canny edge detection algorithm of computer vision, primitive character image is processed, extract life of swimming
The Edge texture feature of thing, i.e. planktonic local feature, obtain local feature image;
S4: build the multiple features fusion convolutional neural networks model based on primitive character, global characteristics and local feature, should
Multiple features fusion convolutional neural networks include three separate basic sub-networks, and each basic sub-network is respectively trained original
Characteristic image, global characteristics image and local feature image, according to practical situation and demand, described basis sub-network can use
Any one convolutional neural networks in alexnet, vggnet or googlenet, based on multiple features fusion convolutional Neural net
The final classification accuracy of network model can step up according to the difference of selected basis sub-network, correspondingly, model training
Time cost also can be stepped up, in the present embodiment, basic sub-network employs alexnet convolutional neural networks, wherein,
1 to 5 layers of this multiple features fusion convolutional neural networks are convolutional layer, and 6 to 8 layers is full articulamentum;
As shown in Fig. 2 each basic sub-network is separate in this model, mutually incoherent, each basic sub-network is
One convolutional neural networks, and structure configuration is identical: each basic sub-network comprises 5 convolutional layers, and each convolutional layer
Convolution kernel taper into, convolution kernel increasing number, each layer of convolution kernel size is respectively 11x11,11x11,5x5,3x3 and
3x3, and convolution kernel quantitative classification is 96,96,384,384 and 256.
It is that multiple Feature Mapping figures are directly merged mostly in prior art, in the present invention, in order that three kinds of features are more
Merge well, fully excavate with high-dimensional and level information, as a kind of preferred technical scheme, in three basic sub-networks
Afterwards, be a full articulamentum mixing together, in described full articulamentum, three bases are merged using the full cross-mixing method that connects
The plinth sub-network Feature Mapping figure that obtains of training, have three layers full articulamentum altogether, is distributed in pyramid shape, every layer complete connect
Number progressively successively decreases, and every layer of full connection neuron number is fixed as 2048.
Differ greatly from view of between global characteristics image and local feature image, this full connection mixing together side
Method, compared with commonly full articulamentum directly merges, effectively reduces global characteristics image and local feature image co-registration is brought
Error, realizes the abundant fusion of multiple features, lifts plankton image classification accuracy rate.
S5: whole primitive character images, global characteristics image and the local that step s1, step s2 and step s3 are obtained
Characteristic image is input in this multiple features fusion convolutional neural networks model and is trained, and finally gives the multiple features after optimization and melts
Conjunction convolutional neural networks model:
S51: initial state information is set first, including iterationses, learning rate and initialization mode;
S52: this multiple features fusion convolutional neural networks model is carried out with fl transmission and feeds back with backward, make this multiple features melt
Close convolutional neural networks model to be trained according to the plankton image of input and learn;
S53: output loss function value and accuracy rate;
S54: lift the performance of this multiple features fusion convolutional neural networks model by reducing loss function value;
S54: judge whether to reach the iterationses of setting, if it is, training finishes, the multiple features after being optimized melt
Close convolutional neural networks model, otherwise, continue to redirect execution step s52;
S6: plankton image to be sorted is input in the multiple features fusion convolutional neural networks model after optimization,
According to the probability score of final output, judge the classification corresponding to plankton image.
This invention, on the premise of based on a large amount of training datas, is up in the classification accuracy of multi-class plankton image
95%.
In above-described embodiment of the application, by providing a kind of plankton based on multiple features fusion convolutional neural networks
Image classification method, gathers a large amount of clearly plankton images first, builds plankton picture number multi-class on a large scale
According to collection, subsequently use image conversion and Boundary extracting algorithm, extract global characteristics and local feature, by primitive character image, entirely
Office's characteristic image is put into together with local feature image in the multiple features fusion convolutional neural networks of deep learning and is trained, and obtains many
The convolutional neural networks model of Feature Fusion, finally, plankton image is input to this multiple features fusion convolutional neural networks
In model, according to the probability score of final output, you can realize classification.The present invention is by biological morphology angle, computer vision
Method is combined with deep learning technology, has very high classification particularly with multi-class plankton image on a large scale accurate
Rate.
It should be pointed out that described above is not limitation of the present invention, the present invention is also not limited to the example above,
Change, modification, interpolation or replacement that those skilled in the art are made in the essential scope of the present invention, also should
Belong to protection scope of the present invention.
Claims (3)
1. a kind of plankton image classification method based on multiple features fusion convolutional neural networks it is characterised in that include as
Lower step:
S1: collection clearly plankton image, builds plankton image data set multi-class on a large scale, wherein, this number
According to the plankton image concentrated as primitive character image;
S2: process primitive character image, extract planktonic global characteristics, obtain global characteristics image, concrete process step
For:
S21: using image segmentation scharr operator, primitive character image is changed, the image after conversion includes global characteristics
And local feature;
S22: remove the local feature in transition diagram picture using bilateral filtering method;
S23: enhancing contrast ratio is projecting the global characteristics in transition diagram picture;
S3: by the canny edge detection algorithm of computer vision, primitive character image is processed, extract planktonic
Edge texture feature, i.e. planktonic local feature, obtain local feature image;
S4: build the multiple features fusion convolutional neural networks model based on primitive character, global characteristics and local feature, how special this is
Levy fusion convolutional neural networks and include three separate basic sub-networks, each basic sub-network is respectively trained primitive character
Image, global characteristics image and local feature image, wherein, 1 to 5 layers of this multiple features fusion convolutional neural networks is convolution
Layer, 6 to 8 layers is full articulamentum;
S5: whole primitive character images, global characteristics image and the local feature that step s1, step s2 and step s3 are obtained
Image is input in this multiple features fusion convolutional neural networks and is trained, and finally gives the god of the multiple features fusion convolution after optimization
Through network model:
S51: initial state information is set first, including iterationses, learning rate and initialization mode;
S52: this multiple features fusion convolutional neural networks model is carried out with fl transmission and feeds back with backward, make this multiple features fusion roll up
Long-pending neural network model is trained according to the plankton image of input and learns;
S53: output loss function value and accuracy rate;
S54: lift the performance of this multiple features fusion convolutional neural networks model by reducing loss function value;
S54: judge whether to reach the iterationses of setting, the multiple features fusion volume if it is, training finishes, after being optimized
Long-pending neural network model, otherwise, continues to redirect execution step s52;
S6: plankton image to be sorted is input in the multiple features fusion convolutional neural networks model after optimization, according to
The probability score of final output, judges the classification corresponding to plankton image.
2. the plankton image classification method based on multiple features fusion convolutional neural networks according to claim 1, its
It is characterised by, described basis sub-network uses any one the convolutional Neural net in alexnet, vggnet or googlenet
Network.
3. the plankton image classification method based on multiple features fusion convolutional neural networks according to claim 1, its
It is characterised by, in described full articulamentum, the feature that three basic sub-network training obtain is merged using the full cross-mixing method that connects
Mapping graph.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610912684.1A CN106372648B (en) | 2016-10-20 | 2016-10-20 | Plankton image classification method based on multi-feature fusion convolutional neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610912684.1A CN106372648B (en) | 2016-10-20 | 2016-10-20 | Plankton image classification method based on multi-feature fusion convolutional neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106372648A true CN106372648A (en) | 2017-02-01 |
CN106372648B CN106372648B (en) | 2020-03-13 |
Family
ID=57895026
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610912684.1A Active CN106372648B (en) | 2016-10-20 | 2016-10-20 | Plankton image classification method based on multi-feature fusion convolutional neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106372648B (en) |
Cited By (36)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106709478A (en) * | 2017-02-22 | 2017-05-24 | 桂林电子科技大学 | Pedestrian image feature classification method and system |
CN106991666A (en) * | 2017-02-24 | 2017-07-28 | 中国科学院合肥物质科学研究院 | A kind of disease geo-radar image recognition methods suitable for many size pictorial informations |
CN107292250A (en) * | 2017-05-31 | 2017-10-24 | 西安科技大学 | A kind of gait recognition method based on deep neural network |
CN107403430A (en) * | 2017-06-15 | 2017-11-28 | 中山大学 | A kind of RGBD image, semantics dividing method |
CN107506786A (en) * | 2017-07-21 | 2017-12-22 | 华中科技大学 | A kind of attributive classification recognition methods based on deep learning |
CN107610141A (en) * | 2017-09-05 | 2018-01-19 | 华南理工大学 | A kind of remote sensing images semantic segmentation method based on deep learning |
CN107610129A (en) * | 2017-08-14 | 2018-01-19 | 四川大学 | A kind of multi-modal nasopharyngeal carcinima joint dividing method based on CNN |
CN107633258A (en) * | 2017-08-21 | 2018-01-26 | 北京精密机电控制设备研究所 | A kind of deep learning identifying system and method based on feed-forward character extraction |
CN108038459A (en) * | 2017-12-20 | 2018-05-15 | 深圳先进技术研究院 | A kind of detection recognition method of aquatic organism, terminal device and storage medium |
CN108154509A (en) * | 2018-01-12 | 2018-06-12 | 平安科技(深圳)有限公司 | Cancer recognition methods, device and storage medium |
CN108171276A (en) * | 2018-01-17 | 2018-06-15 | 百度在线网络技术(北京)有限公司 | For generating the method and apparatus of information |
CN108229341A (en) * | 2017-12-15 | 2018-06-29 | 北京市商汤科技开发有限公司 | Sorting technique and device, electronic equipment, computer storage media, program |
CN108427957A (en) * | 2017-02-15 | 2018-08-21 | 中国科学院深圳先进技术研究院 | image classification method and system |
CN108491880A (en) * | 2018-03-23 | 2018-09-04 | 西安电子科技大学 | Object classification based on neural network and position and orientation estimation method |
CN108805181A (en) * | 2018-05-25 | 2018-11-13 | 深圳大学 | A kind of image classification device and sorting technique based on more disaggregated models |
CN109190695A (en) * | 2018-08-28 | 2019-01-11 | 中国海洋大学 | A kind of fish image classification method based on depth convolutional neural networks |
CN109190640A (en) * | 2018-08-20 | 2019-01-11 | 贵州省生物研究所 | A kind of the intercept type acquisition method and acquisition system of the planktonic organism based on big data |
CN109711343A (en) * | 2018-12-27 | 2019-05-03 | 北京思图场景数据科技服务有限公司 | Behavioral structure method based on the tracking of expression, gesture recognition and expression in the eyes |
CN109871905A (en) * | 2019-03-14 | 2019-06-11 | 同济大学 | A kind of plant leaf identification method based on attention mechanism depth model |
CN109886933A (en) * | 2019-01-25 | 2019-06-14 | 腾讯科技(深圳)有限公司 | A kind of medical image recognition method, apparatus and storage medium |
CN109993201A (en) * | 2019-02-14 | 2019-07-09 | 平安科技(深圳)有限公司 | A kind of image processing method, device and readable storage medium storing program for executing |
CN110188794A (en) * | 2019-04-23 | 2019-08-30 | 深圳大学 | A kind of training method, device, equipment and the storage medium of deep learning model |
CN110287990A (en) * | 2019-05-21 | 2019-09-27 | 山东大学 | Microalgae image classification method, system, equipment and storage medium |
CN110825381A (en) * | 2019-09-29 | 2020-02-21 | 南京大学 | CNN-based bug positioning method combining source code semantics and grammatical features |
CN111274860A (en) * | 2019-11-08 | 2020-06-12 | 杭州安脉盛智能技术有限公司 | Machine vision-based online automatic tobacco leaf grade sorting identification method |
CN111406264A (en) * | 2017-10-27 | 2020-07-10 | 谷歌有限责任公司 | Neural architecture search |
CN111723714A (en) * | 2020-06-10 | 2020-09-29 | 上海商汤智能科技有限公司 | Method, device and medium for identifying authenticity of face image |
CN111899241A (en) * | 2020-07-28 | 2020-11-06 | 华中科技大学 | Quantitative on-line detection method and system for defects of PCB (printed Circuit Board) patches in front of furnace |
CN111898677A (en) * | 2020-07-30 | 2020-11-06 | 大连海事大学 | Plankton automatic detection method based on deep learning |
CN112016574A (en) * | 2020-10-22 | 2020-12-01 | 北京科技大学 | Image classification method based on feature fusion |
CN112069958A (en) * | 2020-08-27 | 2020-12-11 | 广西柳工机械股份有限公司 | Material identification method, device, equipment and storage medium |
CN112488170A (en) * | 2020-11-24 | 2021-03-12 | 杭州电子科技大学 | Multi-feature fusion image classification method based on deep learning |
CN112652032A (en) * | 2021-01-14 | 2021-04-13 | 深圳科亚医疗科技有限公司 | Organ modeling method, image classification device, and storage medium |
CN113205039A (en) * | 2021-04-29 | 2021-08-03 | 广东电网有限责任公司东莞供电局 | Power equipment fault image identification and disaster investigation system and method based on multiple DCNNs |
CN114842510A (en) * | 2022-05-27 | 2022-08-02 | 澜途集思生态科技集团有限公司 | Ecological organism identification method based on ScatchDet algorithm |
US11682111B2 (en) | 2020-03-18 | 2023-06-20 | International Business Machines Corporation | Semi-supervised classification of microorganism |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105488515A (en) * | 2014-09-17 | 2016-04-13 | 富士通株式会社 | Method for training convolutional neural network classifier and image processing device |
CN105825235A (en) * | 2016-03-16 | 2016-08-03 | 博康智能网络科技股份有限公司 | Image identification method based on deep learning of multiple characteristic graphs |
-
2016
- 2016-10-20 CN CN201610912684.1A patent/CN106372648B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105488515A (en) * | 2014-09-17 | 2016-04-13 | 富士通株式会社 | Method for training convolutional neural network classifier and image processing device |
CN105825235A (en) * | 2016-03-16 | 2016-08-03 | 博康智能网络科技股份有限公司 | Image identification method based on deep learning of multiple characteristic graphs |
Non-Patent Citations (2)
Title |
---|
HUSSEIN A.AL-BARAZANCHI ET AL.: "Performance Evaluation of Hybrid CNN for SIPPER Plankton Image Classificaiton", 《2015 THIRD INERNATIONAL CONFERENCE ON IMAGE INFORMATION PROCESSING》 * |
JIALUN DAI ET AL: "ZooplanktoNet: Deep Convolutional Network for Zooplankton Classification", 《IEEE》 * |
Cited By (53)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108427957A (en) * | 2017-02-15 | 2018-08-21 | 中国科学院深圳先进技术研究院 | image classification method and system |
CN108427957B (en) * | 2017-02-15 | 2021-12-21 | 中国科学院深圳先进技术研究院 | Image classification method and system |
CN106709478A (en) * | 2017-02-22 | 2017-05-24 | 桂林电子科技大学 | Pedestrian image feature classification method and system |
CN106991666B (en) * | 2017-02-24 | 2019-06-07 | 中国科学院合肥物质科学研究院 | A kind of disease geo-radar image recognition methods suitable for more size pictorial informations |
CN106991666A (en) * | 2017-02-24 | 2017-07-28 | 中国科学院合肥物质科学研究院 | A kind of disease geo-radar image recognition methods suitable for many size pictorial informations |
CN107292250A (en) * | 2017-05-31 | 2017-10-24 | 西安科技大学 | A kind of gait recognition method based on deep neural network |
CN107403430A (en) * | 2017-06-15 | 2017-11-28 | 中山大学 | A kind of RGBD image, semantics dividing method |
CN107403430B (en) * | 2017-06-15 | 2020-08-07 | 中山大学 | RGBD image semantic segmentation method |
CN107506786B (en) * | 2017-07-21 | 2020-06-02 | 华中科技大学 | Deep learning-based attribute classification identification method |
CN107506786A (en) * | 2017-07-21 | 2017-12-22 | 华中科技大学 | A kind of attributive classification recognition methods based on deep learning |
CN107610129A (en) * | 2017-08-14 | 2018-01-19 | 四川大学 | A kind of multi-modal nasopharyngeal carcinima joint dividing method based on CNN |
CN107633258A (en) * | 2017-08-21 | 2018-01-26 | 北京精密机电控制设备研究所 | A kind of deep learning identifying system and method based on feed-forward character extraction |
CN107633258B (en) * | 2017-08-21 | 2020-04-10 | 北京精密机电控制设备研究所 | Deep learning identification system and method based on feedforward feature extraction |
CN107610141A (en) * | 2017-09-05 | 2018-01-19 | 华南理工大学 | A kind of remote sensing images semantic segmentation method based on deep learning |
CN107610141B (en) * | 2017-09-05 | 2020-04-03 | 华南理工大学 | Remote sensing image semantic segmentation method based on deep learning |
CN111406264A (en) * | 2017-10-27 | 2020-07-10 | 谷歌有限责任公司 | Neural architecture search |
CN108229341A (en) * | 2017-12-15 | 2018-06-29 | 北京市商汤科技开发有限公司 | Sorting technique and device, electronic equipment, computer storage media, program |
CN108038459A (en) * | 2017-12-20 | 2018-05-15 | 深圳先进技术研究院 | A kind of detection recognition method of aquatic organism, terminal device and storage medium |
CN108154509B (en) * | 2018-01-12 | 2022-11-11 | 平安科技(深圳)有限公司 | Cancer identification method, device and storage medium |
CN108154509A (en) * | 2018-01-12 | 2018-06-12 | 平安科技(深圳)有限公司 | Cancer recognition methods, device and storage medium |
CN108171276A (en) * | 2018-01-17 | 2018-06-15 | 百度在线网络技术(北京)有限公司 | For generating the method and apparatus of information |
CN108171276B (en) * | 2018-01-17 | 2019-07-23 | 百度在线网络技术(北京)有限公司 | Method and apparatus for generating information |
CN108491880A (en) * | 2018-03-23 | 2018-09-04 | 西安电子科技大学 | Object classification based on neural network and position and orientation estimation method |
CN108805181A (en) * | 2018-05-25 | 2018-11-13 | 深圳大学 | A kind of image classification device and sorting technique based on more disaggregated models |
CN108805181B (en) * | 2018-05-25 | 2021-11-23 | 深圳大学 | Image classification device and method based on multi-classification model |
CN109190640A (en) * | 2018-08-20 | 2019-01-11 | 贵州省生物研究所 | A kind of the intercept type acquisition method and acquisition system of the planktonic organism based on big data |
CN109190695A (en) * | 2018-08-28 | 2019-01-11 | 中国海洋大学 | A kind of fish image classification method based on depth convolutional neural networks |
CN109190695B (en) * | 2018-08-28 | 2021-08-03 | 中国海洋大学 | Fish image classification method based on deep convolutional neural network |
CN109711343A (en) * | 2018-12-27 | 2019-05-03 | 北京思图场景数据科技服务有限公司 | Behavioral structure method based on the tracking of expression, gesture recognition and expression in the eyes |
CN109886933A (en) * | 2019-01-25 | 2019-06-14 | 腾讯科技(深圳)有限公司 | A kind of medical image recognition method, apparatus and storage medium |
CN109886933B (en) * | 2019-01-25 | 2021-11-02 | 腾讯科技(深圳)有限公司 | Medical image recognition method and device and storage medium |
CN109993201A (en) * | 2019-02-14 | 2019-07-09 | 平安科技(深圳)有限公司 | A kind of image processing method, device and readable storage medium storing program for executing |
WO2020164278A1 (en) * | 2019-02-14 | 2020-08-20 | 平安科技(深圳)有限公司 | Image processing method and device, electronic equipment and readable storage medium |
CN109871905A (en) * | 2019-03-14 | 2019-06-11 | 同济大学 | A kind of plant leaf identification method based on attention mechanism depth model |
CN110188794B (en) * | 2019-04-23 | 2023-02-28 | 深圳大学 | Deep learning model training method, device, equipment and storage medium |
CN110188794A (en) * | 2019-04-23 | 2019-08-30 | 深圳大学 | A kind of training method, device, equipment and the storage medium of deep learning model |
CN110287990A (en) * | 2019-05-21 | 2019-09-27 | 山东大学 | Microalgae image classification method, system, equipment and storage medium |
CN110825381A (en) * | 2019-09-29 | 2020-02-21 | 南京大学 | CNN-based bug positioning method combining source code semantics and grammatical features |
CN111274860A (en) * | 2019-11-08 | 2020-06-12 | 杭州安脉盛智能技术有限公司 | Machine vision-based online automatic tobacco leaf grade sorting identification method |
CN111274860B (en) * | 2019-11-08 | 2023-08-22 | 杭州安脉盛智能技术有限公司 | Recognition method for online automatic tobacco grade sorting based on machine vision |
US11682111B2 (en) | 2020-03-18 | 2023-06-20 | International Business Machines Corporation | Semi-supervised classification of microorganism |
CN111723714A (en) * | 2020-06-10 | 2020-09-29 | 上海商汤智能科技有限公司 | Method, device and medium for identifying authenticity of face image |
CN111723714B (en) * | 2020-06-10 | 2023-11-03 | 上海商汤智能科技有限公司 | Method, device and medium for identifying authenticity of face image |
CN111899241A (en) * | 2020-07-28 | 2020-11-06 | 华中科技大学 | Quantitative on-line detection method and system for defects of PCB (printed Circuit Board) patches in front of furnace |
CN111898677A (en) * | 2020-07-30 | 2020-11-06 | 大连海事大学 | Plankton automatic detection method based on deep learning |
CN112069958A (en) * | 2020-08-27 | 2020-12-11 | 广西柳工机械股份有限公司 | Material identification method, device, equipment and storage medium |
CN112016574A (en) * | 2020-10-22 | 2020-12-01 | 北京科技大学 | Image classification method based on feature fusion |
CN112488170B (en) * | 2020-11-24 | 2024-04-05 | 杭州电子科技大学 | Multi-feature fusion image classification method based on deep learning |
CN112488170A (en) * | 2020-11-24 | 2021-03-12 | 杭州电子科技大学 | Multi-feature fusion image classification method based on deep learning |
CN112652032B (en) * | 2021-01-14 | 2023-05-30 | 深圳科亚医疗科技有限公司 | Modeling method for organ, image classification device, and storage medium |
CN112652032A (en) * | 2021-01-14 | 2021-04-13 | 深圳科亚医疗科技有限公司 | Organ modeling method, image classification device, and storage medium |
CN113205039A (en) * | 2021-04-29 | 2021-08-03 | 广东电网有限责任公司东莞供电局 | Power equipment fault image identification and disaster investigation system and method based on multiple DCNNs |
CN114842510A (en) * | 2022-05-27 | 2022-08-02 | 澜途集思生态科技集团有限公司 | Ecological organism identification method based on ScatchDet algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN106372648B (en) | 2020-03-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106372648A (en) | Multi-feature-fusion-convolutional-neural-network-based plankton image classification method | |
CN105631426B (en) | The method and device of text detection is carried out to picture | |
CN106157319B (en) | The conspicuousness detection method in region and Pixel-level fusion based on convolutional neural networks | |
CN106709924B (en) | Image, semantic dividing method based on depth convolutional neural networks and super-pixel | |
CN105657402B (en) | A kind of depth map restoration methods | |
CN103247036B (en) | Many exposure images fusion method and device | |
CN109800736A (en) | A kind of method for extracting roads based on remote sensing image and deep learning | |
CN105701508B (en) | Global local optimum model and conspicuousness detection algorithm based on multistage convolutional neural networks | |
CN102314614B (en) | Image semantics classification method based on class-shared multiple kernel learning (MKL) | |
CN107808132A (en) | A kind of scene image classification method for merging topic model | |
CN109255334A (en) | Remote sensing image terrain classification method based on deep learning semantic segmentation network | |
CN107610123A (en) | A kind of image aesthetic quality evaluation method based on depth convolutional neural networks | |
CN107657257A (en) | A kind of semantic image dividing method based on multichannel convolutive neutral net | |
CN105894045A (en) | Vehicle type recognition method with deep network model based on spatial pyramid pooling | |
CN106023145A (en) | Remote sensing image segmentation and identification method based on superpixel marking | |
CN105512676A (en) | Food recognition method at intelligent terminal | |
CN104517122A (en) | Image target recognition method based on optimized convolution architecture | |
CN107122796A (en) | A kind of remote sensing image sorting technique based on multiple-limb network integration model | |
CN109285162A (en) | A kind of image, semantic dividing method based on regional area conditional random field models | |
CN106408030A (en) | SAR image classification method based on middle lamella semantic attribute and convolution neural network | |
CN107506792B (en) | Semi-supervised salient object detection method | |
CN104463242B (en) | The multiple features action identification method that feature based is converted and dictionary learns | |
CN107516103A (en) | A kind of image classification method and system | |
CN110059539A (en) | A kind of natural scene text position detection method based on image segmentation | |
CN104063686A (en) | System and method for performing interactive diagnosis on crop leaf segment disease images |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant |