CN111160416B - Aedes mosquito identification method based on support vector machine - Google Patents
Aedes mosquito identification method based on support vector machine Download PDFInfo
- Publication number
- CN111160416B CN111160416B CN201911275811.1A CN201911275811A CN111160416B CN 111160416 B CN111160416 B CN 111160416B CN 201911275811 A CN201911275811 A CN 201911275811A CN 111160416 B CN111160416 B CN 111160416B
- Authority
- CN
- China
- Prior art keywords
- image
- training
- support vector
- vector machine
- mosquito
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an aedes identification method based on a support vector machine. The method comprises a training phase and a testing phase. In the training stage, image preprocessing is firstly carried out, and the size of the collected colored mosquito image is standardized; then, extracting image characteristics, respectively extracting color histogram characteristics and direction gradient histogram characteristics, and fusing in a serial connection mode; and finally, training a support vector machine classifier, and training the sample by adopting a linear kernel support vector machine function. The testing stage is to carry out pretreatment and feature extraction on the mosquito image to be tested according to the training stage in sequence, input the obtained features into a trained model and output the category of the mosquito image. The method is simple and easy to implement, the extracted features have small dependence on the size, direction and visual angle of the image, and the geometric and optical deformation of the image can be well kept unchanged, so that the method has strong robustness.
Description
Technical Field
The invention belongs to the technical field of digital image processing, and relates to an aedes identification method based on a support vector machine, in particular to an identification method of two mosquito images of aedes albopictus and culex pipiens pallens.
Background
Dengue has become the fastest growing mosquito-borne disease in subtropical and tropical countries, it is a viral disease transmitted by dengue virus vectors, the main vector mosquito being aedes albopictus. Dengue has no specific therapeutic or prophylactic vaccine and prevention of dengue epidemics depends on monitoring and is therefore very important for vector control. The existing method for identifying the aedes mainly depends on manual visual identification under a microscope, and a great deal of time is consumed for identification. The method has the advantages of higher time cost and low efficiency, and also brings serious cognitive burden, thus causing higher false recognition rate. Therefore, the mosquito species are automatically identified by adopting the computer, the working efficiency is greatly improved, and the manpower identification burden is reduced.
Disclosure of Invention
The invention aims to provide an aedes identification method based on a support vector machine.
The method comprises a training phase and a testing phase.
The specific method in the training stage is as follows:
collecting mosquito color images, wherein each image only contains one mosquito to form a training set; the training set only contains color images of two types of mosquitoes, namely aedes albopictus and culex pallens.
Step (1), image preprocessing: the mosquito color image size is normalized by a bicubic interpolation method and reduced to M multiplied by N to obtain an RGB color reduced image, wherein M and N are the length and the width of the color reduced image respectively, and M =128,256, N =64,128.
Step (2), image feature extraction:
a. extracting B component from RGB component of color reduced image, counting histogram HOG with P as interval for B component value, obtaining B component histogram feature H B ；P＝4,8,16,32。
b. Extracting histogram directional gradient feature H G The specific method comprises the following steps: firstly, carrying out weighted average on a color reduced image according to R, G and B three channels to obtain a gray level image, carrying out gray level normalization on the image after gray level is carried out by adopting a Gamma correction method, calculating the gradient of each pixel, then dividing the image into small cells which are not overlapped with each other, and counting the gradient on each small cell by taking Q as an intervalHistogram of degree, every 2 x 2 cells form a block, the HOG characteristics of the block are obtained by concatenating the characteristics of all cells in a block, the width of a cell is overlapped between adjacent blocks, and the HOG characteristics of all blocks in the image are concatenated to obtain the HOG characteristics H of the color reduced image G ；Q∈{20°,30°,36°,45°,60°}。
c. Obtaining the mosquito color reduction image characteristics: for feature H B And feature H G Connected in series to obtain characteristic H = (H) B ,H G )。
Step (3), training a support vector machine classifier: and acquiring the characteristic H of each image in the training set, wherein the label of the Aedes albopictus is 1, the label of the Culex pipiens pallens is-1, and training by adopting a linear kernel support vector machine function.
The specific method in the test stage is as follows:
and (3) sequentially preprocessing the test image according to the step (1) and extracting image characteristics in the step (2), inputting the obtained characteristics H into the model trained in the step (3), wherein the output of the model is 1, which indicates that the test sample is aedes albopictus, and-1, which indicates that the test sample is culex pipiens pallens.
The invention has the beneficial effects that:
the invention provides an aedes identification method based on a support vector machine, which is simple and easy to implement. The extracted color features have small dependence on the size, direction and visual angle of the image, and the extracted HOG features can keep good invariance to the geometric and optical deformation of the image, so that the method has strong robustness.
Detailed Description
The present invention will be described in detail below with reference to examples to provide those skilled in the art with a better understanding of the present invention.
And (3) a flow based on support vector machine aedes identification. The method specifically comprises a training stage and a testing stage:
the specific method in the training stage is as follows:
collecting mosquito color images, wherein each image only contains one mosquito to form a training set; the training set only contains color images of two types of mosquitoes, namely aedes albopictus and culex pallens.
Step (1), image preprocessing: the dimension of the mosquito color image is normalized by a bicubic interpolation method and is reduced to 128 multiplied by 64, and an RGB (red, green and blue) color reduced image is obtained.
Step (2), image feature extraction:
a. extracting B component from RGB component of color reduced image, and counting histogram HOG with interval of P =4 for B component value to obtain B component histogram feature H B Dimension 64.
b. Extracting Histogram of Oriented Gradient (HOG) features H G . Firstly, carrying out weighted average on a color reduced image according to three channels of R, G and B to obtain a gray level image, carrying out gray level normalization on the gray level image by adopting a Gamma correction method, and calculating the gradient of each pixel; the image is then divided into small cells, each cell being 8 x 8 pixels, which do not overlap with each other. Gradient direction angle value is ∈ [0,180 °]The gradient histogram is counted at intervals of Q =20 °, and a vector having a dimension of 9 is obtained. Then every 2 x 2 cells are combined into a block, and the feature vectors of all the cells in the block are connected in series to obtain the HOG feature of the block; overlapping a cell width between adjacent blocks, and obtaining 3780-dimensional histogram feature H of direction gradient for each 128 × 64 scaled picture G 。
c. Obtaining the mosquito color reduction image characteristics: for feature H B And feature H G Connected in series to obtain characteristic H = (H) B ,H G ). Where H has a dimension of 3844.
Step (3), training a support vector machine classifier: and acquiring the characteristic H of each image in the training set, wherein the label of the Aedes albopictus is 1, the label of the Culex pipiens pallens is-1, and training by adopting a linear kernel support vector machine function.
The specific method in the test stage is as follows:
and (3) sequentially preprocessing the test image according to the step (1) and extracting image characteristics in the step (2), inputting the obtained characteristics H into the model trained in the step (3), wherein the output of the model is 1, which indicates that the test sample is aedes albopictus, and-1, which indicates that the test sample is culex pipiens pallens.
Claims (1)
1. The support vector machine-based aedes identification method comprises a training stage and a testing stage, and is characterized in that:
the specific method of the training stage is as follows:
collecting mosquito color images, wherein each image only contains one mosquito to form a training set; the training set only contains color images of Aedes albopictus and Culex pipiens pallens;
step (1), image preprocessing: normalizing the size of the mosquito color image by a bicubic interpolation method, and reducing the mosquito color image to M multiplied by N to obtain an RGB color reduced image, wherein M and N are the length and the width of the color reduced image respectively, and M =128,256, N =64,128;
step (2), image feature extraction:
a. extracting B component from RGB component of color reduced image, counting histogram HOG of B component value with P as interval to obtain B component histogram feature H B ；P＝4,8,16,32；
b. Extracting histogram directional gradient feature H G The specific method comprises the following steps: firstly, carrying out weighted average on a color reduced image according to R, G and B three channels to obtain a gray level image, carrying out gray level normalization on the image after gray level is carried out by adopting a Gamma correction method, calculating the gradient of each pixel, then dividing the image into small cells which are not overlapped with each other, counting a gradient histogram on each small cell by taking Q as an interval, forming a block by every 2 x 2 cells, connecting the characteristics of all the cells in one block in series to obtain the HOG characteristic of the block, overlapping the width of one cell between adjacent blocks, and connecting the HOG characteristics of all the blocks in the image in series to obtain the HOG characteristic H of the color reduced image G ；Q∈{200,300,360,450,600}；
c. Obtaining the mosquito color reduction image characteristics: for feature H B And feature H G Connected in series to obtain the characteristic H = (H) B ,H G )；
Step (3), training a support vector machine classifier: acquiring the characteristic H of each image in a training set, wherein the label of Aedes albopictus is 1, the label of Culex pipiens pallens is-1, training is carried out by adopting a support vector machine, and the kernel function of the support vector machine adopts a linear kernel function;
the specific method of the test stage is as follows:
and (3) sequentially preprocessing the test image according to the step (1) and extracting image characteristics in the step (2), inputting the obtained characteristics H into the model trained in the step (3), wherein the output of the model is 1, which indicates that the test sample is aedes albopictus, and-1, which indicates that the test sample is culex pipiens pallens.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911275811.1A CN111160416B (en) | 2019-12-12 | 2019-12-12 | Aedes mosquito identification method based on support vector machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911275811.1A CN111160416B (en) | 2019-12-12 | 2019-12-12 | Aedes mosquito identification method based on support vector machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111160416A CN111160416A (en) | 2020-05-15 |
| CN111160416B true CN111160416B (en) | 2023-03-21 |
Family
ID=70557015
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911275811.1A Active CN111160416B (en) | 2019-12-12 | 2019-12-12 | Aedes mosquito identification method based on support vector machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111160416B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113505815A (en) * | 2021-06-15 | 2021-10-15 | 杭州电子科技大学 | Mosquito identification method based on multi-feature fusion |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104077580A (en) * | 2014-07-15 | 2014-10-01 | 中国科学院合肥物质科学研究院 | Pest image automatic recognition method based on high-reliability network |
| CN105447859A (en) * | 2015-11-18 | 2016-03-30 | 扬州大学 | Field wheat aphid counting method |
| CN108537154A (en) * | 2018-03-28 | 2018-09-14 | 天津大学 | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning |
| CN109034269A (en) * | 2018-08-22 | 2018-12-18 | 华北水利水电大学 | A kind of bollworm female male imago method of discrimination based on computer vision technique |
| CN110516686A (en) * | 2019-07-11 | 2019-11-29 | 西安电子科技大学 | The mosquito recognition methods of three color RGB images |
-
2019
- 2019-12-12 CN CN201911275811.1A patent/CN111160416B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104077580A (en) * | 2014-07-15 | 2014-10-01 | 中国科学院合肥物质科学研究院 | Pest image automatic recognition method based on high-reliability network |
| CN105447859A (en) * | 2015-11-18 | 2016-03-30 | 扬州大学 | Field wheat aphid counting method |
| CN108537154A (en) * | 2018-03-28 | 2018-09-14 | 天津大学 | Transmission line of electricity Bird's Nest recognition methods based on HOG features and machine learning |
| CN109034269A (en) * | 2018-08-22 | 2018-12-18 | 华北水利水电大学 | A kind of bollworm female male imago method of discrimination based on computer vision technique |
| CN110516686A (en) * | 2019-07-11 | 2019-11-29 | 西安电子科技大学 | The mosquito recognition methods of three color RGB images |
Non-Patent Citations (3)
| Title |
|---|
| 基于区域生长和SVM结合的黄金大蚊快速检测算法;林探宇等;《广东农业科学》;20160710(第07期);172-177 * |
| 基于多特征融合和稀疏表示的农业害虫图像识别方法;张永玲等;《中国农业科学》;20180601(第11期);全文 * |
| 基于韦伯局部描述子和颜色直方图的鳞翅目昆虫翅图像特征描述与种类识别;竺乐庆等;《昆虫学报》;20150420(第04期);419-426 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111160416A (en) | 2020-05-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ranjan et al. | Detection and classification of leaf disease using artificial neural network | |
| Husin et al. | Feasibility study on plant chili disease detection using image processing techniques | |
| CN109344883A (en) | Fruit tree diseases and pests recognition methods under a kind of complex background based on empty convolution | |
| CN105447859A (en) | Field wheat aphid counting method | |
| Padol et al. | Fusion classification technique used to detect downy and Powdery Mildew grape leaf diseases | |
| Islam et al. | Rice leaf disease recognition using local threshold based segmentation and deep CNN | |
| CN112597495B (en) | Malicious code detection method, system, equipment and storage medium | |
| CN109146878A (en) | A kind of method for detecting impurities based on image procossing | |
| CN102637262B (en) | A kind of self-adaptation bacterial counting | |
| CN107767379A (en) | Pcb board marks print quality inspection method | |
| Masood et al. | Plants disease segmentation using image processing | |
| CN105005969B (en) | A kind of bill images alter detection method and system | |
| CN111160416B (en) | Aedes mosquito identification method based on support vector machine | |
| CN105512612A (en) | SVM-based image classification method for capsule endoscope | |
| CN108711160B (en) | Target segmentation method based on HSI (high speed input/output) enhanced model | |
| CN110348342A (en) | A kind of piping disease image partition method based on full convolutional network | |
| CN109961416A (en) | A kind of business license information extracting method based on Morphological Gradient Multiscale Fusion | |
| CN109325498A (en) | The Vein extraction algorithm of Canny operator is improved based on window dynamic threshold | |
| CN107958253A (en) | A kind of method and apparatus of image recognition | |
| CN110455201A (en) | Stalk plant height measurement method based on machine vision | |
| CN115187980A (en) | Calcein AM/PI double-staining cell fluorescence image-based cell counting method | |
| CN110569716A (en) | Goods shelf image copying detection method | |
| Sannakki et al. | SVM-DSD: SVM based diagnostic system for the detection of pomegranate leaf diseases | |
| CN109087287A (en) | Method based on image analysis evaluation space cabin material surface bacterial plaque elimination effect | |
| CN112164055A (en) | Photovoltaic cell color difference classification method based on color segmentation |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |