CN102239793B - Real-time classification method and system of rice pests - Google Patents
Real-time classification method and system of rice pests Download PDFInfo
- Publication number
- CN102239793B CN102239793B CN201110095670A CN201110095670A CN102239793B CN 102239793 B CN102239793 B CN 102239793B CN 201110095670 A CN201110095670 A CN 201110095670A CN 201110095670 A CN201110095670 A CN 201110095670A CN 102239793 B CN102239793 B CN 102239793B
- Authority
- CN
- China
- Prior art keywords
- image
- test sample
- images
- training
- rice
- 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.)
- Expired - Fee Related
Links
- 235000007164 Oryza sativa Nutrition 0.000 title claims abstract description 22
- 235000009566 rice Nutrition 0.000 title claims abstract description 22
- 238000000034 method Methods 0.000 title claims abstract description 18
- 241000607479 Yersinia pestis Species 0.000 title abstract description 12
- 240000007594 Oryza sativa Species 0.000 title 1
- 238000012549 training Methods 0.000 claims abstract description 22
- 241000209094 Oryza Species 0.000 claims abstract description 21
- 238000012360 testing method Methods 0.000 claims abstract description 18
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 9
- 239000011159 matrix material Substances 0.000 claims description 11
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000001514 detection method Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 abstract description 3
- 238000005286 illumination Methods 0.000 abstract description 2
- 239000000284 extract Substances 0.000 abstract 1
- 230000009466 transformation Effects 0.000 abstract 1
- 241000238631 Hexapoda Species 0.000 description 24
- 230000005540 biological transmission Effects 0.000 description 5
- 241001414720 Cicadellidae Species 0.000 description 3
- 241000008892 Cnaphalocrocis patnalis Species 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 241000665051 Chilo sacchariphagus indicus Species 0.000 description 2
- 241000426497 Chilo suppressalis Species 0.000 description 2
- 241001498622 Cixius wagneri Species 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 2
- 241001556089 Nilaparvata lugens Species 0.000 description 2
- 241000176086 Sogatella furcifera Species 0.000 description 2
- 235000021329 brown rice Nutrition 0.000 description 2
- 235000013339 cereals Nutrition 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 229920000742 Cotton Polymers 0.000 description 1
- 241000086608 Empoasca vitis Species 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 230000009418 agronomic effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 201000007094 prostatitis Diseases 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Images
Landscapes
- Catching Or Destruction (AREA)
Abstract
The invention discloses a real-time classification method and system of rice pests. The method disclosed by the invention comprises sample training and sample detecting, wherein the sample training comprises the following steps: (1) collecting multiple training samples of various rice pests and obtaining the images of the rice pests, (2) carrying out the gray transformation on the images, and (3) extracting the characteristic values of the images by adopting a compressed sensing algorithm; and the sample detecting comprises the following steps: obtaining the images of a test sample, processing the images of the test sample according to the steps (2) and (3), obtaining the characteristic values of the images of the test sample, and determining the type of the test sample by utilizing the most adjacent algorithm. The method disclosed by the invention extracts the characteristic values of the images by adopting the compressed sensing algorithm, effectively inhabits the noise, has no requirements on the angle and illumination intensity of photography and is suitable for the real-time identification of pests in fields.
Description
Technical field
The present invention relates to the plant protection technology field, relate in particular to a kind of rice grub real-time grading method and system.
Background technology
Diseases and pests of agronomic crop is one of main agricultural disaster of China, and it has, and kind is many, influence is big, also break out the characteristics of causing disaster often, and its occurrence scope and the order of severity often cause heavy losses to Chinese national economy, particularly agricultural production.According to data; The whole world has at least 6% crops eaten by insect in vegetative period every year, if human and material resources and technology do not catch up with, may reach 10%-30%; And in China because insect pest; The whole nation annual grain and cotton loss about 15%, the main cause that causes this situation be exactly pest detection that the forecast is inaccurate is true, untimely, opportunity is missed in control.
Paddy rice is one of world's staple food crop.China's rice growing area accounts for 1/4 of national cereal crops, and output then accounts for over half.Zhejiang Province's paddy rice cultivated area and output also occupy domestic prostatitis.So, be very to be necessary to the research of rice insect pest.
Conventional artificial detects common employing dish and claps, lures methods such as collection, utilizes artificial sense to check insect at the scene, by instrument or the direct kinds of with the naked eye differentiating insect such as magnifying glass, microscopes, and statistical magnitude.This method is a kind of the most directly perceived, easy but very extensive method, and the investigation work amount is big, and single detection area coverage is little, and efficient is lower, and the investigation cost is high.
Summary of the invention
The invention provides a kind of rice grub real-time grading method, this method has solved that the conventional method workload is big, and efficient is low, the problem that cost is high.
A kind of rice grub real-time grading method comprises sample training and pattern detection:
Described sample training comprises:
(1) training sample of the various rice grubs of collection is some, obtains image;
(2) image is carried out gradation conversion; Need the background removal of image be obtained the image of insect before the conversion; The formula of gradation conversion is y=0.3r+0.59g+0.11b, and wherein: y representes the gray value behind the gray processing, the value of r presentation video red component; The value of g presentation video green component, the value of b presentation video blue component;
(3) after the gradation conversion, adopt the eigen value of compressed sensing algorithm abstract image;
With picture size i * h is example (i is a row, h for row), with the gray value of each picture point be that element makes up gray matrix, fold with windrow, make it become the vectorial p of a row (p ∈ R
1 * n, n=i * h), p and compressed sensing observing matrix at random
Eigen value (vector) x that multiplies each other and obtain image,
Be the gaussian random matrix, n representes line number, and m representes columns, and m＜＜n.
(4) described pattern detection comprises:
Obtain the image of test sample book, handle the image of test sample book (2), (3) set by step, obtains the eigen value of the image of test sample book, utilizes the most contiguous algorithm, confirms the type of test sample book.
The most contiguous algorithm comprises: the distance between the eigen value of calculating test sample image and the eigen value of training sample image; If a class pest comprises a plurality of training samples; Then all distances are asked average; Find the minimum training sample of average distance, the insect type of this training sample is the insect type of test sample book.The formula that calculates distance between two eigen values is following:
The present invention also provides a kind of rice grub real-time grading system, comprising:
Camera is clapped and is got the image of tested insect;
The video decode module is with image decoding back input processor;
Memory module, memory image and sample training result;
Processor is classified to tested insect according to the sample training result;
Display screen shows classification results.
Preferably, said system also comprises wireless transport module.
The beneficial effect that the present invention has is:
(1) realized the real-time grading of rice grub,, the insect picture has been real-time transmitted to far-end, supplied the plant protection expert further to add up, file and segment through wireless transmission;
(2) the compressed sensing theory application is obtained to the insect image feature value, adopted random matrix to extract eigen value, not only effectively suppressed noise, and angle and the intensity of illumination of taking pictures are not done requirement, suitable more field Real time identification insect;
(3) the compressed sensing theory application obtained to the insect image feature value, greatly reduce the dimension of data simultaneously, be adapted at conditional lower computer system such as storage, speed and use.
The present invention is mainly theoretical with compressed sensing, the nearest neighbour classification algorithm combines with advanced data processor, 3G wireless transmission method; Design portable rice grub system for rapidly identifying; Can make things convenient for the farmland management personnel to prevent and treat fast to the rice grub classification.
Description of drawings
Fig. 1 is the structural representation of system of the present invention;
Fig. 2 is the concrete image that divides class pest among the embodiment;
Fig. 3 is the operational flowchart of system of the present invention.
Embodiment
Rice grub real-time grading system as shown in Figure 1; The memory that comprises processor (CPU) 1 and the video decode module 2 that is connected with processor 1, form by program storage and data storage 3, two control buttons 4, four indicator lamps 5, display screen 6 and wireless transport modules 8; The input of video decode module 2 connects camera 7, and power module 9 is supplied power to said apparatus.
The said system workflow is as shown in Figure 2, and after the power-on module 9, indicator lamp A lights; The voltage of expression whole system is normal; The CPU of native system adopts the dsp processor of ADI company, after system powers on, and the program after cpu reset is accomplished in the loading procedure memory; Loading procedure rear indicator light B lights, and CPU is in proper working order in expression.
Camera 7 can be taken the insect image in real time, and the analog video signal that video decode module 2 is exported camera 7 is decoded into and meets the data signal input processor 1 ITU656 standard agreement, the YUV4:2:2 form.The A that pushes button, CPU gathers image and simultaneously the image of gathering is stored in the data storage (SDRAM), simultaneously image is discerned processing, insect is classified, and classification results is shown on display screen 6.
System adopts wireless transport module 8 to realize wireless transmission; Wireless transport module 8 links to each other with CPU through USB interface; The indication CPU that pushes button behind the B sends and gathers good image, and CPU at first connects with far-end server through wireless transport module, carries out the image transmission after connecting; Adopt the TCP host-host protocol, light indicator lamp D after sub-picture transmission is accomplished.
With image shown in Figure 3 (150 * 105 pixel) is example, and the method for said apparatus identification insect is:
1, at first the background of image is scratched and remove; Then image being carried out gradation conversion handles; The formula that gradation conversion is handled is y=0.3r+0.59g+0.11b, and wherein: y representes the gray value after the gradation conversion, the value of r presentation video red component; The value of g presentation video green component, the value of b presentation video blue component;
2, the gray value with each picture point of image after the gradation conversion is the gray matrix that element makes up the insect image, and is folded with windrow to the element of this gray matrix, obtains the vectorial p of row (p ∈ R
1 * n, n=15750), p and random gaussian matrix
Multiplying each other obtains the eigen value x of insect image,
Before detection, also need carry out sample training; Each 20 of the striped rice borer, paddy stem borer, rice leaf roller, pink rice borer, small brown rice planthopper, white-backed planthopper, brown planthopper, Da Bai leafhopper, green leaf hopper adult sample image of gathering different attitudes are as training sample; Handle according to step 1,2 then, obtain the eigen value of all insect images.
3, according to the most contiguous algorithm insect is classified, detailed process is:
Calculate through following formula between the eigen value of eigen value and each training sample image of test sample book apart from χ
2,
The mean value of distance is respectively 2499753.5,267186.9,3929645.3,2687734.8,2811369.3,4316368.1,3793615.3,4516837.5 and 3086334.5 between the eigen value of the eigen value of insect image shown in Figure 2 and pink rice borer, rice leaf roller, big clear leafhopper, striped rice borer, paddy stem borer, small brown rice planthopper, white-backed planthopper, brown planthopper, each 20 sub-picture of Da Bai leafhopper; Can judge that this insect is a rice leaf roller; Be consistent with actual conditions, visible system of the present invention is practicable.
Claims (1)
1. rice grub real-time grading method comprises sample training and pattern detection:
Described sample training comprises:
(1) training sample of the various rice grubs of collection is some, obtains their image;
(2) background of removal image is carried out gradation conversion to image;
(3) eigen value of employing compressed sensing algorithm abstract image:
With each gray values of pixel points of image is that element makes up gray matrix, folded with windrow, makes it become a vectorial p of row, p ∈ R
1 * n, n=i * h, i and h are respectively the line number and the columns of gray matrix; Vectorial p of row and compressed sensing observing matrix at random
Multiplying each other obtains the eigen value x of image, x ∈ R
1 * m,
Be the gaussian random matrix, n representes line number, and m representes columns, and m<<n;
Described pattern detection comprises:
Obtain the image of test sample book, handle the image of test sample book (2), (3) set by step, obtains the eigen value of the image of test sample book, utilizes the most contiguous algorithm, confirms the type of test sample book.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110095670A CN102239793B (en) | 2011-04-18 | 2011-04-18 | Real-time classification method and system of rice pests |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110095670A CN102239793B (en) | 2011-04-18 | 2011-04-18 | Real-time classification method and system of rice pests |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102239793A CN102239793A (en) | 2011-11-16 |
| CN102239793B true CN102239793B (en) | 2012-10-10 |
Family
ID=44958186
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110095670A Expired - Fee Related CN102239793B (en) | 2011-04-18 | 2011-04-18 | Real-time classification method and system of rice pests |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102239793B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103345634B (en) * | 2013-07-29 | 2016-12-28 | 湖南省植物保护研究所 | The automatic identifying method of Vegetables insecticide on a kind of yellow plate |
| CN104918007B (en) * | 2015-05-15 | 2017-10-27 | 浙江大学 | Crop field insect pest situation monitoring sampling apparatus and the method for sampling based on computer vision |
| CN106054844B (en) * | 2016-07-12 | 2019-04-30 | 河海大学 | A kind of agricultural intelligent remote management system |
| CN106203453B (en) * | 2016-07-18 | 2019-05-28 | 清华大学深圳研究生院 | A kind of compressed sensing based biological and abiotic target identification method and its system |
| CN112508012B (en) * | 2020-12-01 | 2022-09-06 | 浙大宁波理工学院 | Orchard pest intelligent positioning and identifying method suitable for small target sample |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101766106A (en) * | 2009-12-31 | 2010-07-07 | 上海量科电子科技有限公司 | Plant-nursing terminal with prompt function, system and system implementation method |
| CN202084060U (en) * | 2011-04-18 | 2011-12-21 | 浙江大学 | Real-time classifying system of paddy rice pest |
-
2011
- 2011-04-18 CN CN201110095670A patent/CN102239793B/en not_active Expired - Fee Related
Non-Patent Citations (1)
| Title |
|---|
| 杨玲玲.水稻飞虱自动识别技术的研究.《南京农业大学硕士学位论文》.2008, * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102239793A (en) | 2011-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fernandez-Gallego et al. | Wheat ear counting in-field conditions: high throughput and low-cost approach using RGB images | |
| US11048940B2 (en) | Recognition of weed in a natural environment | |
| Chen et al. | A YOLOv3-based computer vision system for identification of tea buds and the picking point | |
| Nagai et al. | Using digital camera images to detect canopy condition of deciduous broad-leaved trees | |
| CN102239793B (en) | Real-time classification method and system of rice pests | |
| Rasti et al. | A survey of high resolution image processing techniques for cereal crop growth monitoring | |
| Zhou et al. | Qualification of soybean responses to flooding stress using UAV-based imagery and deep learning | |
| Kelly et al. | An opinion on imaging challenges in phenotyping field crops | |
| Beniaich et al. | Determination of vegetation cover index under different soil management systems of cover plants by using an unmanned aerial vehicle with an onboard digital photographic camera | |
| Hu et al. | Estimation of leaf chlorophyll content of rice using image color analysis | |
| Zang et al. | Detection method of wheat spike improved YOLOv5s based on the attention mechanism | |
| JPWO2012073519A1 (en) | Leaf area index measurement system, apparatus, method and program | |
| Zhang et al. | High-throughput corn ear screening method based on two-pathway convolutional neural network | |
| Özreçberoğlu et al. | Mathematical models for the estimation of leaf chlorophyll content based on RGB colours of contact imaging with smartphones: A pomegranate example | |
| Yang et al. | Cotton hail disaster classification based on drone multispectral images at the flowering and boll stage | |
| Li et al. | An automatic approach for detecting seedlings per hill of machine-transplanted hybrid rice utilizing machine vision | |
| Sunoj et al. | Digital image analysis estimates of biomass, carbon, and nitrogen uptake of winter cereal cover crops | |
| Song et al. | Assessment of wheat chlorophyll content by the multiple linear regression of leaf image features | |
| CN202084060U (en) | Real-time classifying system of paddy rice pest | |
| Bleasdale et al. | Feasibility of detecting apple scab infections using low-cost sensors and interpreting radiation interactions with scab lesions | |
| Li et al. | Variety identification of delinted cottonseeds based on BP neural network | |
| CN112098415B (en) | Nondestructive testing method for quality of waxberries | |
| Guan et al. | A corn canopy organs detection method based on improved DBi-YOLOv8 network | |
| Chen et al. | An efficient approach to monitoring pine wilt disease severity based on random sampling plots and UAV imagery | |
| Mehrubeoglu et al. | Detecting red blotch disease in grape leaves using hyperspectral imaging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20121010 Termination date: 20210418 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |