CN110742047B - Automatic paper changing system based on computer vision effect and insect pest monitoring system - Google Patents

Automatic paper changing system based on computer vision effect and insect pest monitoring system Download PDF

Info

Publication number
CN110742047B
CN110742047B CN201910965923.3A CN201910965923A CN110742047B CN 110742047 B CN110742047 B CN 110742047B CN 201910965923 A CN201910965923 A CN 201910965923A CN 110742047 B CN110742047 B CN 110742047B
Authority
CN
China
Prior art keywords
paper
image
layer
pest
insect
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
Application number
CN201910965923.3A
Other languages
Chinese (zh)
Other versions
CN110742047A (en
Inventor
黄仁杰
姚廷山
郑勇强
黄伟
赵亦欣
李涛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southwest University
Original Assignee
Southwest University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Southwest University filed Critical Southwest University
Priority to CN201910965923.3A priority Critical patent/CN110742047B/en
Publication of CN110742047A publication Critical patent/CN110742047A/en
Application granted granted Critical
Publication of CN110742047B publication Critical patent/CN110742047B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
    • A01M—CATCHING, TRAPPING OR SCARING OF ANIMALS; APPARATUS FOR THE DESTRUCTION OF NOXIOUS ANIMALS OR NOXIOUS PLANTS
    • A01M1/00—Stationary means for catching or killing insects
    • A01M1/14—Catching by adhesive surfaces
    • A01M1/16—Fly papers or ribbons
    • A—HUMAN NECESSITIES
    • A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
    • A01M—CATCHING, TRAPPING OR SCARING OF ANIMALS; APPARATUS FOR THE DESTRUCTION OF NOXIOUS ANIMALS OR NOXIOUS PLANTS
    • A01M1/00—Stationary means for catching or killing insects
    • A01M1/02—Stationary means for catching or killing insects with devices or substances, e.g. food, pheronones attracting the insects
    • A01M1/026—Stationary means for catching or killing insects with devices or substances, e.g. food, pheronones attracting the insects combined with devices for monitoring insect presence, e.g. termites
    • 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
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
    • 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]
    • 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
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30181—Earth observation
    • G06T2207/30188—Vegetation; Agriculture

Landscapes

  • Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Pest Control & Pesticides (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Environmental Sciences (AREA)
  • Zoology (AREA)
  • Wood Science & Technology (AREA)
  • Geometry (AREA)
  • Insects & Arthropods (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Quality & Reliability (AREA)
  • Artificial Intelligence (AREA)
  • Catching Or Destruction (AREA)

Abstract

The invention discloses an automatic paper changing system and a pest monitoring system based on computer vision effect. The automatic paper changing system comprises film-covered pest sticking paper, a paper changing device, an image acquisition device and a processor; the tectorial membrane mythimna separata paper sets up in trading the paper device, and treater control trades the paper device and changes the mythimna separata paper. The insect sticking paper is of a double-layer (including a working layer and a protective layer) structure, the working layer comprises a substrate layer internally soaked with attractant, a sticky layer provided with breathable micropores and a solid attractant layer, and the protective layer is stuck to the surface of the working layer. The paper changing device adopts a double-shaft linkage mode to change new sticky trap paper and recover used sticky trap paper. The insect pest monitoring system counts specific target insects based on the deep neural network and collects environmental data at the same time. The invention can realize long-term, effective and continuous trapping of pests and acquisition of pest image data, and can also identify and count pest species and quantity by adopting a deep neural network technology, thereby accurately monitoring pest scale and trend.

Description

Automatic paper changing system based on computer vision effect and insect pest monitoring system
Technical Field
The invention relates to the field of insect pest monitoring and early warning, in particular to an automatic insect sticking paper replacing system based on a computer visual effect and an insect pest monitoring system based on the automatic insect sticking paper replacing system.
Background
Disease and pest monitoring and management are important production links in agricultural production. The deployment of a pest disaster monitoring and early warning system is an important means for monitoring and managing pests in agricultural production management departments and large modern farms. Pest disaster monitoring and early warning system mainly includes: pest data collection, pest data analysis, statistics and trend prediction.
The pest data collection in a long-term and timely manner is the most basic link of the pest disaster monitoring and early warning system. Most of the methods adopt a mode of 'distributed fixed-point collection and centralized summary' to collect pest data in an area. The long-term collection of target pests and breeding environment data thereof is helpful for forming target pest big data and developing a more reliable early warning and prediction system. The pest population distribution and the propagation trend in the area can be mastered in time by collecting pest data in time, and early warning is realized in time. However, since the pests are distributed in a specific area, a specific crop, or a specific part of a fruit tree, either without, with, or scattered, it is very difficult to collect the basic data of the pests continuously, effectively, for a long period of time.
Currently, most pest monitoring systems employ manual or semi-automated methods for pest trapping to collect and analyze pest data. Of these, the "trapper + computer vision system" is the most popular monitoring approach. For example, the trapping device is used for trapping pests of trypanosomatidae such as bactrocera minax, bactrocera dorsalis and bactrocera cucurbitae, and moth pests such as prodenia litura, spodoptera exigua and cabbage looper, and then the pests are collected, reported and identified by a manual or computer vision system to form pest basic data.
However, most current traps are oriented toward trapping pests rather than monitoring pests. For example, although insects may be trapped, visual systems may be configured to monitor the images taken, such as conventional ball traps with wicks, rectangular sticky boards, rectangular sticky paper, McPhail type traps, and the like. However, these traps are not suitable for either being equipped with an automatic attractant change device or for storing the attractant in large quantities at the installation site, necessitating manual attractant change for long-term monitoring purposes.
CN103598169A discloses a fruit fly monitoring device and system based on McPhail type trap, wherein a liquid attractant is contained in a trap bottle, a camera is installed at the bottle mouth to collect the pest image captured by the liquid, the device must periodically and manually replace the attractant in the trap bottle to keep the trap bottle working effectively for a long time, and fruit flies are trapped in the liquid, insects can move in the liquid or are immersed in the liquid, so that the fuzzy or distinguishable characteristics of the collected pest image are changed, thereby seriously affecting the later identification statistics.
CN104186449A discloses a pest monitoring device and system based on a sticky trap type trap, which employs a method of attracting and separating a sex attractant from a sticky trap, places the sex attractant above the sticky trap, and employs multiple sets of motor transmission systems to replace and recover the corresponding sticky traps. Because the attractant is separated from the pest sticking plate, the trapper can be automatically replaced, the attractant cannot be automatically replaced, and the attractant is easy to rapidly lose efficacy after being completely exposed, cannot work effectively for a long time, and must be manually replaced to keep the insects to be effectively and continuously trapped and monitored. Further, it employs a separate manner of sex attraction and color attraction, and each attraction manner can achieve only a single attraction effect, cannot combine to enhance the attraction effect, and ignores the food attraction, and thus the attraction effect is not substantially enhanced, and the effectiveness of the attraction cannot be enhanced.
CN208609748U discloses a pest sticking device for pest and disease monitoring system, the mythimna separata paper structure that wherein adopts is: viscose is laid on one side of the pest sticking paper, a protective film is pasted on the viscose, the film scraping device is assembled on the fixed frame body, and the film scraping device is used for scraping the protective film. The mode can play a role in protecting unused pest sticking paper. However, on one hand, the scraped protective film is wasted, the scraped protective film cannot be effectively treated, and long-term continuous use is affected, on the other hand, the effectiveness of the viscose and the paper changing area can be ensured only by strictly controlling the film scraping amount, and meanwhile, the trapping effect of the sticky paper is not improved, so that the trapping effect is not strong.
CN105454182A proposes a citrus fruit fly trap with replaceable insect-sticking plate, which has four hanging columns for hanging four insect-sticking plates at the same time, but the design purpose is to facilitate manual replacement of the insect-sticking plates. It must be changed through manual, can't realize the automatic change of mythimna separata board.
In summary, the attractants, traps, replacement devices, trapping devices and monitoring methods of the existing devices and systems cannot be organically coordinated to realize long-term, effective and continuous pest data acquisition and pest identification and statistics without manual maintenance. Most of the existing attractant has the problems of quick invalidation, poor trapping effect, manual replacement of the trapping agent, poor control of the replacement precision or the effectiveness of the trap by automatic replacement effect, incapability of effectively acquiring pest data for a long time, inaccurate pest statistical information caused by low pest automatic classification precision and the like.
Disclosure of Invention
The invention aims to: in order to solve the existing problems, the automatic pest sticking paper replacing system based on the computer vision effect is provided. The pest sticking paper is stored in a large quantity, so that the pest sticking paper is automatically replaced, the pest sticking paper is ensured to have longer service life and long-time working effectiveness, pests are trapped and pest data are collected effectively and continuously for a long time, and manual maintenance is not needed. A pest monitoring system is also provided. The method and the device can accurately monitor specific target pests and provide effective data support for pest early warning.
The technical scheme adopted by the invention is as follows:
an automatic change system of tectorial membrane mythimna separata paper based on computer vision effect, it includes:
the film-coated pest sticking paper is arranged on the paper changing device and comprises a first film layer and a second film layer, the first film layer comprises a substrate layer, an adhesive layer and a solid attractant layer, the substrate layer, the adhesive layer and the solid attractant layer are sequentially arranged to form a working surface of the first film layer, the second film layer is a non-breathable film layer, and the second film layer is attached to the working surface of the first film layer; the base layer is a paper-impregnated layer impregnated with a liquid attractant, the surface of the base layer, which is far away from the adhesive layer, is subjected to air permeation prevention treatment (namely the exposed surface of the base layer is subjected to air permeation prevention treatment), the adhesive layer is coated on the surface of the base layer, the adhesive layer is provided with air-permeable micropores, and the solid attractant layer is at least arranged on the surface of the adhesive layer;
the paper changing device is used for storing unused tectorial membrane sticky paper in a winding mode, and when a paper changing signal is received, the used sticky paper is changed by the stored unused tectorial membrane sticky paper, and the paper changing device comprises: outputting the unused film-coated insect-sticking paper, and separating a first film layer and a second film layer of the unused film-coated insect-sticking paper in the output process to expose the first film layer; and synchronously recycling the used first film layer and the second film layer in a winding mode in the process of outputting the film-coated insect-sticking paper. The paper changing process mainly comprises a separation and recovery process. Which includes separating the first and second film layers of unused coated sticker, exposing the first film layer sticker to a trap pest (disposed in a position to or replacing the first film layer that has been used), and enabling the second film layer to be effectively recycled. The first film layer and the second film layer of the film-coated pest sticking paper are wound to one end part to be recombined for synchronous recovery in the recovery process, so that effective storage and saving of storage space are realized.
The image acquisition device is used for at least acquiring an image of the working area of the sticky paper as an image of the working surface of the sticky paper;
and the processor is respectively connected with the paper changing device and the image acquisition device and is used for sending a paper changing signal to the paper changing device when the insect density in the image of the working surface of the sticky paper reaches a preset value or reaches a preset paper changing time.
In the system, the pest sticking paper adopts a double-layer film structure, and a large amount of pest sticking paper can be stored in a winding mode. Meanwhile, the double-layer film structure can effectively prevent the insect sticking paper from losing efficacy after being stored for a long time. The cooperation of the liquid attractant and the solid attractant can greatly improve the attraction effect. The design of the ventilation micropores enables the liquid attractant to be slowly volatilized, and further the service life of the pest sticking paper is greatly prolonged. The system can automatically replace the sticky paper in time based on the monitoring of time or the density of the sticky paper insects, and has strong timeliness and high automation degree. The insect sticking paper can automatically work for a long time under the condition of stable power supply after being manually supplemented once by matching with the insect sticking paper of the invention.
Further, both sides of the first film layer are subjected to sealing treatment. The two sides are sealed, so that the effective components of the attractant can be prevented from being volatilized, and the viscose part is oxidized, so that the pest sticking paper is ensured not to lose efficacy after being stored for a long time.
Further, the solid attractant layer is composed of solid attractant particles embedded in the adhesive layer. The granular solid attractant can influence the adhesion of the adhesive layer to insects as little as possible.
Further, trade the paper device and include the shell and install the paper mechanism that trades in the shell, trade the paper mechanism and include first spool, the second spool, first spool group and second spool group are divided equally and are do not contained a plurality of rollers, first spool is used for the storage mythimna separata paper, first rete that draws forth from first spool is through first spool group back, be connected to the second spool, second rete that draws forth from first spool is through second spool group back, be connected to the second spool, the second spool rotates first rete and the second rete in order to retrieve the mythimna separata paper in step, the shell is provided with uncovered, uncovered will expose outside through whole or part of the first rete of first spool group.
When the sticky note paper needs to be replaced, the second scroll is rotated, the used sticky note paper starts to be recycled, the first roller is further driven to rotate so as to output the unused sticky note paper, in the process of recycling the sticky note paper by the second scroll, the first film layer (sticky note paper) and the second film layer (sealing film) of the film-coated sticky note paper output by the first scroll are separated at the adjacent end parts of the first scroll group and the second scroll group, the two film layers move towards two different directions respectively, and finally the two film layers are recycled by the second roller. And when the newly output first film layer is completely filled with the opening, stopping rotating the second reel. At this point, the replacement of the first membrane layer is completed. By adopting the device, the pest sticking paper is separated by a certain amount, so that waste is avoided. The mode is in a simple mode, the pest sticking paper is replaced rapidly (the pest sticking paper is recycled and new pest sticking paper is deployed). The paper changing device is simple in structure and high in space utilization rate, and can store a large amount of pest sticking paper. Except the open part, the rest parts are all sealed by the shell, which can play a better protection role for the internal facilities.
Further, the uncovered second spool side tip that is close to of shell is provided with the clearance and scrapes, and this clearance is scraped and is contacted with first rete. This clearance is scraped when retrieving the armyworm paper, can realize being glued the clearance of big individual insect, and little individual insect can be retrieved along with the armyworm paper after by two adjacent roller bearing roll-in.
Further, the path length of the first roller set is not greater than the path length of the second roller set. First rete and second rete need constitute the annular, design like this, and the annular arc is filled to the use second rete as far as possible, and then the first rete of separation (working face) as far as possible, prevents that the mythimna separata paper effective part from being wasted because of filling annular arc, improves the utilization ratio to the mythimna separata paper.
Further, the paper changing signal comprises a paper changing starting signal and a paper changing stopping signal, and the processor sends the paper changing starting signal to the paper changing device when the insect density in the image of the working surface of the sticky paper reaches a preset value or when the preset paper changing time is reached so as to control the paper changing device to start to change the paper; and detecting the image acquired by the image acquisition device, calculating the difference between the acquired image and the image of the blank sticky paper, and sending a paper change stopping signal to the paper change device when the difference is smaller than a preset difference threshold value so as to control the paper change device to finish paper change. This design can realize the accurate control to trading the paper process, prevents that the mythimna separata paper from being too much separated or the separation is not enough.
A pest damage monitoring system based on above-mentioned automatic change system of mythimna separata paper based on computer vision effect, it still includes:
the environment data acquisition device is used for acquiring environment data;
the wireless communication module is used for carrying out wireless communication with a far end;
the processor is also respectively connected with the environmental data acquisition device and the wireless communication module and used for processing and identifying the sticky paper working face images, classifying and respectively counting the insects in the sticky paper working face images, and sending the sticky paper working face images, the insect types and the corresponding quantity as well as the corresponding environmental data to the far end through the wireless communication module.
The invention simultaneously collects insect data such as insect images, types and numbers and corresponding environment data such as temperature, humidity, illumination and the like, forms more complete structural pest data, and can better utilize the pest data to estimate pest disaster trends.
Further, the method for processing the image of the working face of the sticky paper by the processor comprises the following steps:
A. calculating a difference image between the collected working surface image of the sticky paper and the last collected working surface image of the sticky paper, and performing gray processing on the difference image to obtain a gray image; removing the detected and segmented image to highlight the newly captured insects;
B. carrying out binarization operation on the gray level image to obtain a binarization image;
C. performing morphological processing on the binary image;
D. carrying out contour detection on the image processed by the C, and calculating the center and the area of each contour region;
E. filtering each contour region in the binary image according to preset upper and lower limit area thresholds so as to eliminate non-target pests; the area threshold value can be set according to the size of the target pest body;
F. and corresponding to each contour region left after filtering, cutting a subgraph from the collected working surface image of the sticky paper to obtain the insect to be identified. And marking the corresponding positions in the image by corresponding the rest contour regions to the positions in the image of the working surface of the pest sticking paper, wherein the marked regions are obtained by dividing.
The method can realize the screening of the effective monitoring targets, automatically filter the ineffective targets and ensure the pertinence to the monitoring targets. By setting different thresholds, the monitoring target can be flexibly changed.
Further, the method for identifying the insects to be identified by the processor comprises the following steps:
A. utilizing an insect construction database similar to the preset target pests to fine tune the pre-trained ResNet50 model, then utilizing an image database of the preset target pests to fine tune the coarse-tuned ResNet50 model, and taking the final ResNet50 as a feature extractor;
B. training a plurality of classes of classifiers based on a support vector machine by using an image database of the preset target pest, wherein the feature vectors of the image database of the preset target pest are extracted by ResNet 50;
C. and extracting a characteristic vector of the insect to be identified in the sticky paper working face image by adopting ResNet50, and identifying by utilizing a trained classifier.
The method realizes automatic identification of target pests based on the pest species identification of the deep convolutional neural network, greatly reduces the manual identification cost, and greatly improves the accuracy of pest identification. Meanwhile, the method can effectively utilize the existing non-target pest data, improve the distinguishability of the learned characteristics and solve the problem of insufficient target pest training data.
In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:
1. the insect sticking paper adopted by the invention is convenient for on-site mass storage. The attractant greatly enhances the attraction effect by adopting a mode of combining various attractants. The structure of the working layer enables the liquid attractant to be slowly released, and further, when the insect sticking paper is used, the service cycle of the attractant is prolonged, namely, the quality guarantee period of the insect sticking paper is prolonged.
2. According to the invention, by designing the novel pest sticking paper and matching with the double-shaft linkage paper replacing device, the automatic replacement of the pest sticking paper is realized based on the computer vision recognition technology, and the replacement timeliness is high. By means of winding (aiming at the pest sticking paper designed by the invention), a large amount of pest sticking paper can be stored in the device, and then the pest sticking paper is replaced fully automatically after being supplemented manually, so that the manual maintenance cost is reduced, and the working continuity of monitoring pests based on the pest sticking paper is improved.
3. The invention has pertinence to the monitoring of insect pests, can automatically filter invalid targets, and can realize the accurate positioning and identification of insects based on the insect pest identification technology of the neural network. Thereby ensuring the effectiveness of insect pest monitoring.
4. The invention integrates environmental data and can ensure effective support of collected data on insect pest early warning by matching with an accurate monitoring target. Compared with the traditional full-scale statistics of insects, the method has stronger pertinence and data integrity and richness.
5. When the pest sticking paper is arranged, the efficient utilization of the pest sticking paper can be realized, and the waste is avoided.
Drawings
The invention will now be described, by way of example, with reference to the accompanying drawings, in which:
fig. 1 is one embodiment of a pest monitoring system configuration.
Fig. 2 is one embodiment of an embedded device configuration in a pest monitoring system.
Fig. 3 is a flowchart of the paper change control.
FIG. 4 is a flow chart of a method of insect localization and segmentation.
Fig. 5 is a flow chart of a method of insect identification.
Fig. 6 is a structural view of the sheet changer.
Fig. 7 is a side view of the internal facility of fig. 6.
Fig. 8 is a structural view of fig. 7 without the second roller set.
Fig. 9 is a structural view of fig. 7 without the first roller group.
Fig. 10 is a structural view of the pest sticking paper.
Fig. 11 is a sectional view taken along the length of fig. 10.
Detailed Description
All of the features disclosed in this specification, or all of the steps in any method or process so disclosed, may be combined in any combination, except combinations of features and/or steps that are mutually exclusive.
Any feature disclosed in this specification (including any accompanying claims, abstract) may be replaced by alternative features serving equivalent or similar purposes, unless expressly stated otherwise. That is, unless expressly stated otherwise, each feature is only an example of a generic series of equivalent or similar features.
Example one
The embodiment discloses automatic change system of tectorial membrane mythimna separata paper based on computer vision effect includes:
the film-covered pest sticking paper is arranged on the paper changing device, and the working area of the film-covered pest sticking paper is used for trapping insects. As shown in fig. 10 and 11, the pest paper is a double-layer pest paper structure, and includes a first film layer 9 and a second film layer 10, where the first film layer 9 includes a substrate layer 91, an adhesive layer 92, and a solid attractant layer 94, the substrate layer 91, the adhesive layer 92, and the solid attractant layer 94 are sequentially arranged to form a working surface of the first film layer 9, the second film layer 10 is a non-air-permeable film layer, and the second film layer 10 is attached to the working surface of the first film layer 9; the base layer 91 is a paper-impregnated layer impregnated with the liquid attractant, the surface 910 of the base layer 91, which is away from the adhesive layer 92, is subjected to air permeation prevention treatment, the adhesive layer 92 is coated on the surface of the base layer 91, the adhesive layer 92 is provided with air permeation micropores 93, and the solid attractant layer 94 is at least arranged on the surface of the adhesive layer 92. Bilayer structure's mythimna separata paper, on the one hand, is sealed with the non-ventilative rete on the working layer, can prevent that the effective attractant in the mythimna separata paper that is not unpacked from volatilizees the inefficacy, prevents that the adhesive oxidation from becoming invalid, and then reaches the effect of depositing for a long time, on the other hand, adopts the rete structural design, also can realize the coiling storage of mythimna separata paper, is convenient for a large amount of mythimna separata papers of single storage. The liquid and solid attractants in the insect sticking paper can be designed specifically according to different application scenes (for trapping different insects). Typically, the liquid attractant is a sex attractant or a fruit-flavored attractant, and the solid attractant layer 94 is a food attractant, so as to achieve the effect of separating and acting together the various attractants. For example, for citrus fruit flies, the liquid attractant includes methyl eugenol and/or ethyl acetate. The solid attractant layer 94 is comprised of solid attractant particles that are uniformly sprinkled over the exposed surface of the adhesive layer. The air-permeable pores 93 of the adhesive layer 92 are arranged in units of columns along the length of the vermin-sticking paper in one embodiment. In order to prevent the solid attractant particles from blocking the air-permeable micro-holes 93 to influence the volatilization of the liquid attractant, the pore diameter of the air-permeable micro-holes 93 is larger than the diameter of the solid attractant particles. Considering that the first film layer 9 of the insect-sticking paper is composed of a plurality of sub-layers which are not integrally molded, an interface exists between the layers, and if the insect-sticking paper is stored for a long period of time, the interface may be slowly oxidized or the effective component may be volatilized. In this case, the first film layer 9 is sealed at both side edges. That is, both side edges of the sub-layers (the base layer 91 and the adhesive layer 92, and the solid attractant layer 94 is sprinkled on the adhesive layer, and thus, the sub-layers are sealed by the second film layer 10) of the first film layer 9 are sealed. The seal herein may be a paraffin or other greasy material seal. Usually, some insects are sensitive to color, a certain attracting effect can be achieved by adopting common colored paper, in the insect sticking paper, the adhesive layer 92 is a transparent layer, and the substrate layer 91 is designed to be yellow, so that the insect sticking paper has a certain attracting effect on citrus fruit flies.
The paper changing device is used for storing unused tectorial membrane sticky paper in a winding mode and replacing the used tectorial membrane sticky paper by using the stored unused tectorial membrane sticky paper to a working area when receiving a paper changing signal, and comprises: outputting the unused laminated pest sticking paper, and separating a first film layer and a second film layer of the unused laminated pest sticking paper in the output process to replace the used first film layer with the first film layer, wherein a new first film layer is exposed; in the process of outputting the film-coated insect-sticking paper, the used first film layer and the previously separated second film layer are synchronously recovered in a winding mode. The length of the recovered first film layer and the second film layer corresponds to the length of the output film-covered pest sticking paper.
The image acquisition device is arranged right opposite to the working face of the pest sticking paper of the pest sticking device and is used for at least acquiring the image of the working area of the pest sticking paper as the image of the working face of the pest sticking paper.
And the processor is respectively connected with the paper changing device and the image acquisition device and is used for processing and identifying the image of the working surface of the sticky paper, and the insect density in the image of the working surface of the sticky paper reaches a preset value or sends a paper changing signal to the paper changing device when the preset paper changing time is reached.
The method for calculating the insect density r comprises the following steps:
Figure BDA0002230473660000111
wherein, PtAll pixels, P, of the rectangular yellow sticker image of the finger (sticker work area) trapiRefers to the number of pixels occupied by the detected ith insect.
Specifically, referring to fig. 3, the strategy flow of sending the paper change signal by the processor is as follows:
1. detecting a timer signal and calculating insect density;
2. when the timing of the timer signal is judged to reach the threshold value of the validity period of the attractant or the insect density is judged to reach the preset threshold value, a paper changing signal is sent out to trigger the paper changing device to change the paper;
the control of the paper changing device on the sticky paper can be step control, that is, the sticky paper with a preset length (area) is changed each time, or the control is carried out through computer vision recognition, and the specific steps are as follows:
referring to fig. 3, the processor detects an image acquired by the image acquisition device, respectively extracts LBP (local binarization pattern) features of the acquired image and a preset background image (an image corresponding to new sticky paper), calculates an euclidean distance (feature difference) between the image and the preset background image features, determines that the used sticky paper is completely replaced when the distance is less than a difference threshold, and at this time, the processor sends a paper replacement stopping signal to the paper replacement device to control the paper replacement device to stop replacing the sticky paper. After the paper is replaced, the paper replacement time needs to be updated so as to determine the shelf life of the pest sticking paper again.
Example two (this part you should read carefully, against the reference numerals, some of which are wrong or confusing)
The present embodiment discloses the structure of a paper changer. As shown in fig. 6 and 7, the paper changing device includes a housing 1 and a paper changing mechanism disposed in the housing 1, the paper changing mechanism includes a first reel 2, a second reel 3, a first reel group 4 and a second reel group 5, the first reel group 4 and the second reel group 5 each include a plurality of rollers, the first reel 2 stores unused double-layer coated insect-sticking paper by winding, a protective layer 7 (i.e., a second film layer) of the insect-sticking paper stored in the first reel 2 passes through (each roller of) the second reel group 5 and is fixed to the second reel 3, a working layer (i.e., a first film layer) 8 of the insect-sticking paper passes through (each roller of) the first reel group 4 and is fixed to the second reel 3 and is finally recovered to the second reel 3, the housing 1 is provided with an opening 14, and the opening 14 exposes all or part of the working layer 8 passing through the first reel group 4. The working layer 8 is used as a working surface of the pest sticking paper for sticking the pests, so that the pest sticking paper needs to be exposed outside. The exposed part of the opening 14 is the working area of the sticky paper. The working layer 8 exposed by the opening 14 is planar so as to collect images of trapped pests for further processing and identification.
After the package coated insect sticker is installed on the first reel 2, the protective layer 7 and the working layer 8 at the end part of the free end of the insect sticker are separated, the protective layer 7 is fixed to the second reel 3 after passing through the second reel group 5, and the working layer 8 is wound on the second reel after passing through the first reel group 4. The second reel rotates, the protective layer 7 and the used working layer 8 can be synchronously recycled, and the pest sticking paper can be synchronously output on the first reel 2 to supplement the working area, so that the purpose of replacing the exposed working layer 8 is achieved.
The rollers at the two ends of the first roller group 4 and the second roller group 5 are in close proximity, and as shown in fig. 8 and 9, the rollers 41 and 42 are the rollers at the two ends of the first roller group 4, the rollers 51 and 52 are the rollers at the two ends of the second roller group 5, the rollers 41 and 51 form one pair of rollers, and the rollers 42 and 52 form the other pair of rollers. This makes the protective layer 7 and the working layer 8 of the pest sticking paper separate near the first reel 2 end, and the protective layer 7 and the working layer 8 are attached and recovered near the second reel 3 end again. In the case that the working layer 8 is recovered after being cleaned, the pests which are not completely cleaned can be rolled and recovered by two adjacent rollers at the end of the second reel 3.
In order to reduce the volume of the paper changing device as much as possible when the exposed area of the working layer 8 is kept constant, the first roller set 4 and the second roller set 5 are arranged on the outer sides of the first reel 2 and the second reel 3 and are in a mutually separated structure to surround the first reel 2 and the second reel 3. In one embodiment, the housing 1 is in a flat elliptic cylinder shape, i.e., the cross section perpendicular to the axial direction is in an elliptic shape (race track shape), and the first reel 2 and the second reel 3 are mounted in the housing 1 in an up-and-down structure. Specifically, the housing 1 includes a bottom plate 11 for fixing the internal facilities, a cylinder 12 for accommodating the internal facilities, and a cover plate 13, the cylinder 12 is connected to the bottom plate 11 and the cover plate 13 at both ends, the bottom plate 11 is formed in an elliptical shape, the first reel 2 and the second reel 3 are mounted on the long axis of the elliptical bottom plate 11, and further, the two reel mounting positions are concentric with the arc-shaped edges of both ends, respectively. The cylinder 12 is fixed on the bottom plate 11, and the cover plate 13 is detachably connected on the cylinder 12. The axes of the rollers of the first scroll 2, the second scroll 3, the first roller group 4 and the second roller group 5 are all parallel to the axis of the shell 1, and the scrolls and the rollers are all installed in the shell 1 in a plugging manner. As shown in fig. 3, the first roller group 4 and the second roller group 5 form a quadrangle, such as a hexagon, the first roller group 4 includes 3 rollers, the first roller group 4 includes 5 rollers, the second roller group 5 includes 5 rollers, the first roller group 4 and the second roller group 5 are respectively provided with a roller at two adjacent ends, the second roller group 5 forms a quadrangle, the first roller group 4 forms two remaining sides of the hexagon, that is, the path length of the second roller group 5 is not less than that of the first roller group 4, so as to separate the working layer 8 as little as possible, and the protective layer 7 is used to fill up the remaining ring-shaped portion. Utilizing the second; wherein two adjacent rollers of the first roller set 4 near the first roller 2 support the working layer 8 to be exposed to the opening 14 of the housing 1, specifically, the two rollers are disposed at two ends of the opening 14. In order to further improve the supporting effect of the exposed working layer 8, a roller is further arranged in the middle of the opening 14 (i.e. between two rollers supporting the working layer 8 exposed in the opening 14 of the housing 1).
The second reel 3 is driven to rotate by a low-speed motor. The motor can be controlled by a simple switching value. In one embodiment, the motor is controlled (e.g., replaced at regular time) by a single chip microcomputer and a micro-control switch, and the output end of the motor is connected to the rotating shaft of the second winding shaft 3 through a reduction gear. When the motor is started, the second scroll 3 drives the first scroll 3 to simultaneously change paper, and when the motor is stopped, the friction between the protective layer 7 and the working layer 8 and between the roller group 4 and the roller group 5 stops the scroll from rotating.
The shell 1 is provided with a deinsectization mechanism so as to clean individual large insects on the used working layer 8 when the second reel 3 recovers the sticky insect paper, thereby facilitating the recovery of the sticky insect paper. In one embodiment, the insect repelling mechanism is a cleaning scraper 6 disposed at the end of the opening 14 of the housing 1 near the second reel 3.
EXAMPLE III
The embodiment discloses a pest monitoring system of an automatic pest sticking paper replacing system based on the computer vision effect in the first embodiment, and the pest monitoring system further comprises an environmental data acquisition device, a pest control device and a pest control device, wherein the environmental data acquisition device is used for acquiring environmental data; such as the temperature and humidity of the air and soil, the data acquisition can be realized through corresponding sensors. And the wireless communication module is used for carrying out wireless communication with a far end. The processor is also respectively connected with the environmental data acquisition device and the wireless communication module and used for classifying and respectively counting the insects in the sticky paper working face image when the sticky paper working face image is processed and identified, and the sticky paper working face image, the correspondingly identified insect types and quantity and the corresponding environmental data are sent to the far end through the wireless communication module. So as to facilitate the remote analysis of insect pests by combining a plurality of data. In one embodiment, the processor is respectively connected with the environment data acquisition device and the wireless communication module, the environment data acquisition device acquires environment data such as temperature, humidity and illumination of the environment, and the processor forms complete structured pest data by the environment data, the pest images obtained by detection and segmentation, the classification of the pest images and the corresponding counting result, and then transmits the complete structured pest data to the control management center in a distributed manner through the wireless module.
Example four
In one embodiment, as shown in fig. 2, the pest monitoring system is constructed using embedded devices. The embedded device is mainly composed of an NVIDIA Jetson Nano microcomputer (i.e. a processor), and comprises a 4-core arm A57, a 128-core Maxewell architecture GPU and a 4-gigabit memory. The camera (namely an image acquisition device) is connected with the Jetson Nano through a CSI interface. Jetson Nano is connected with a micro-control switch through an I/O port, and the micro-control switch is connected with a motor (namely a paper changing device) to control the change of the pest sticking paper. The temperature and humidity sensor (namely an environmental data acquisition device) is connected to an I/O port of Jetson Nano through an A/D converter. The 4G communication module (SIM7600CE-L, i.e. wireless communication module) is responsible for wireless communication. The Jetson Nano is powered by a 24V direct current power supply, and the rest modules are powered by the Jetson Nano.
Corresponding to the paper changer in the second embodiment described above, the motor is connected to the second reel shaft through a reduction gear to control the rotation of the second reel shaft.
The embodiment also discloses a mounting structure of the insect pest monitoring system. As shown in FIG. 1, the mounting frame D is mounted on the ground, and is respectively provided with a camera B and a paper changer C, the camera B is designed to be opposite to an opening C2 of the paper changer C, and the embedded device A is also mounted on the mounting frame D. Embedded device A is connected to trades paper device C's motor C1 on the mounting bracket D, still installs antenna E, temperature sensor F and humidity transducer G, and the three all is connected to embedded device A, and wherein, humidity transducer stretches into the ground along mounting bracket D at the bottom of, and temperature sensor F and antenna E's probe stretch out embedded device A's shell.
EXAMPLE five
Before the insects in the working surface image of the sticky paper are identified, the insects need to be segmented out. The embodiment discloses an insect detection method for segmenting an image of a working face of sticky insect paper. In this embodiment, the method is mainly used for detecting Bactrocera minax, Bactrocera dorsalis, Bactrocera mandarina and other insects which are likely to be caught by the trap, such as flies, bees and apis cerana and the like, and which have similar body sizes. Other body size insects are excluded, for example, insects of particularly large or small body size. With I(i-1)Representing the last acquired image of the working surface of the sticker, I(i)Representing the currently acquired image, as shown in fig. 4, the insect detection method comprises the steps of:
1. calculation of I(i-1))And I(i)The difference image is subjected to gray level processing to be converted into a difference gray level image;
2. performing OTSU binarization operation on the difference gray level image to obtain a corresponding binarization image;
3. performing erosion and dilation morphological operations on the binarized map;
4. performing contour detection and calculating the center and area of the contour;
5. deleting the too small candidate region and the too large candidate region according to the area size; judging whether the area is too large or too small by using a set upper limit area threshold and a set lower limit area threshold;
6. and marking targets in the working surface image of the pest sticking paper by using a rectangular frame corresponding to the rest candidate areas, and segmenting the targets by using the rectangular frame as a segmentation object.
EXAMPLE six
Continuing, this example discloses a method of identifying the type of insects that have been separated.
Corresponding to the method for detecting insects in the fifth embodiment, the identification method of the present embodiment is mainly used for identifying three insects, namely bactrocera minax, bactrocera dorsalis and bactrocera satsumae, and the identification methods of other insects are the same. The method for identifying insects, as shown in fig. 5, comprises the following steps:
1. collecting images of insects and target pests similar to fruit flies from databases disclosed by iNaturalist, ImageNet, IP102 and the like to construct a large database, finely adjusting a pre-trained ResNet50 model, enabling a trained feature extractor to better distinguish confusing insects, and finally taking the trained ResNet50 as the feature extractor.
2. And training a bactrocera minax, bactrocera dorsalis, bactrocera mandarina and other insect 4-class classifiers by using the collected bactrocera minax image database based on the ResNet50 deep convolution network feature extractor and a support vector machine.
3. And identifying the target pests to be identified detected from the image of the working surface of the sticky insect paper by adopting a trained classifier.
Correspondingly, for the counting of insect species, after the insect species is identified in step 3 above, 1 is added to the corresponding insect species statistical data.
The invention is not limited to the foregoing embodiments. The invention extends to any novel feature or any novel combination of features disclosed in this specification and any novel method or process steps or any novel combination of features disclosed.

Claims (9)

1. An automatic change system of tectorial membrane mythimna separata paper based on computer vision effect, characterized by, includes:
the film-coated pest sticking paper is arranged on the paper changing device and comprises a first film layer and a second film layer, the first film layer comprises a substrate layer, an adhesive layer and a solid attractant layer, the substrate layer, the adhesive layer and the solid attractant layer are sequentially arranged to form a working surface of the first film layer, the second film layer is a non-breathable film layer, and the second film layer is attached to the working surface of the first film layer; the base layer is a paper-impregnated layer impregnated with a liquid attractant, the surface of the base layer, which is far away from the adhesive layer, is subjected to anti-ventilation treatment, the adhesive layer is coated on the surface of the base layer, the adhesive layer is provided with ventilation micropores, and the solid attractant layer is at least arranged on the surface of the adhesive layer;
trade the paper device for through the tectorial membrane mythimna separata paper that the winding mode storage is not used, when receiving and trade the paper signal, utilize the tectorial membrane mythimna separata paper that stores to change the mythimna separata paper that has used, include: outputting the unused film-coated insect-sticking paper, and separating a first film layer and a second film layer of the unused film-coated insect-sticking paper in the output process to expose the first film layer; in the process of outputting the laminated pest sticking paper, the used first film layer and the second film layer are synchronously recycled in a winding mode;
the image acquisition device is used for at least acquiring an image of the working area of the sticky paper as an image of the working surface of the sticky paper;
the processor is respectively connected with the paper changing device and the image acquisition device and is used for sending a paper changing signal to the paper changing device when the insect density in the image of the working surface of the sticky paper reaches a preset value or reaches a preset paper changing time; the processor sends a paper changing starting signal to the paper changing device when the insect density in the image of the working surface of the sticky paper reaches a preset value or when preset paper changing time is reached so as to control the paper changing device to start to change paper; and detecting the image acquired by the image acquisition device, calculating the difference between the acquired image and the image of the blank sticky paper, and sending a paper change stopping signal to the paper change device when the difference is smaller than a preset difference threshold value so as to control the paper change device to finish paper change.
2. The system for automatically replacing film-covered insect-sticking paper as claimed in claim 1, wherein both sides of the first film layer are sealed.
3. The system for automatically replacing film-coated insect-sticking paper as claimed in claim 1 or 2, wherein the solid attractant layer is composed of solid attractant particles embedded in the adhesive layer.
4. The automatic film-covered insect-sticking paper replacing system as claimed in claim 1, wherein the paper replacing device comprises a housing and a paper replacing mechanism installed in the housing, the paper replacing mechanism comprises a first roller, a second roller, a first roller set and a second roller set, the first roller set and the second roller set respectively comprise a plurality of rollers, the first roller is used for storing the insect-sticking paper, the first film layer led out from the first roller passes through the first roller set and is connected to the second roller, the second film layer led out from the first roller passes through the second roller set and is connected to the second roller, the second roller rotates to synchronously recycle the first film layer and the second film layer of the insect-sticking paper, the housing is provided with an opening, and the opening exposes all or part of the first film layer passing through the first roller set.
5. The system for automatically replacing film-covered pest sticking paper as claimed in claim 4, wherein a cleaning scraper is arranged at the end of the opening of the shell close to the end of the second reel.
6. The system for automatically replacing film-covered pest sticking paper as claimed in claim 4 or 5, wherein the path length of the first roller group is not greater than the path length of the second roller group.
7. A pest monitoring system based on the automatic replacement system of the coated sticky paper based on the computer vision effect according to any one of claims 1 to 6, further comprising:
the environment data acquisition device is used for acquiring environment data;
the wireless communication module is used for carrying out wireless communication with a far end;
the processor is further respectively connected with the environmental data acquisition device and the wireless communication module and used for processing and identifying the sticky trap paper working face images, classifying and respectively counting insects in the sticky trap paper working face images, and sending the sticky trap paper working face images, the insect types and the corresponding quantity to the remote end through the wireless communication module and sending the corresponding environmental data to the remote end.
8. The pest monitoring system of claim 7 wherein the processor processes the sticker work surface image by:
A. calculating a difference image between the collected working surface image of the sticky paper and the last collected working surface image of the sticky paper, and performing gray processing on the difference image to obtain a gray image;
B. carrying out binarization operation on the gray level image to obtain a binarization image;
C. performing morphological processing on the binary image;
D. carrying out contour detection on the binary image subjected to the morphological processing, and calculating the center and the area of each contour region;
E. filtering each contour region in the binary image after the morphological processing according to a preset upper limit area threshold and a preset lower limit area threshold;
F. and corresponding to each contour region left after filtering, cutting a subgraph from the collected working surface image of the sticky paper to obtain the insect to be identified.
9. A pest monitoring system according to claim 8 wherein the processor identifies the insect to be identified includes:
A. roughly adjusting a pre-trained ResNet50 model by utilizing an insect image construction database similar to a preset target pest, then finely adjusting the roughly adjusted ResNet50 model by utilizing an image database of the target pest, and taking the final ResNet50 model as a feature extractor;
B. training a plurality of classes of classifiers based on a support vector machine by utilizing a pest image database of the preset target, wherein the feature vector of the pest image database of the preset target is extracted by a refined ResNet50 model;
C. and extracting a characteristic vector of the insect to be identified in the working face image of the sticky paper by adopting a finely adjusted ResNet50 model, and identifying by using a trained classifier.
CN201910965923.3A 2019-10-12 2019-10-12 Automatic paper changing system based on computer vision effect and insect pest monitoring system Active CN110742047B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910965923.3A CN110742047B (en) 2019-10-12 2019-10-12 Automatic paper changing system based on computer vision effect and insect pest monitoring system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910965923.3A CN110742047B (en) 2019-10-12 2019-10-12 Automatic paper changing system based on computer vision effect and insect pest monitoring system

Publications (2)

Publication Number Publication Date
CN110742047A CN110742047A (en) 2020-02-04
CN110742047B true CN110742047B (en) 2021-12-17

Family

ID=69278096

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910965923.3A Active CN110742047B (en) 2019-10-12 2019-10-12 Automatic paper changing system based on computer vision effect and insect pest monitoring system

Country Status (1)

Country Link
CN (1) CN110742047B (en)

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111165455A (en) * 2020-02-20 2020-05-19 桂林电子科技大学 Crop disease and pest supervision device based on machine vision
CN111700045A (en) * 2020-05-11 2020-09-25 五邑大学 Intelligent insect catching device based on image recognition, control method and storage medium
CN112243971A (en) * 2020-11-12 2021-01-22 广东海洋大学 Spodoptera frugiperda trapping monitoring system
CN113678804B (en) * 2021-08-20 2022-07-22 浙江大学 Intelligent tea garden insect attracting plate replacing device and method based on image channel calculation
CN216292718U (en) * 2021-09-14 2022-04-15 深圳卫捕士科技有限公司 Insect catching lamp
CN114304105B (en) * 2021-12-10 2023-07-21 青阳县青宁农业科技发展有限责任公司 Intelligent trapping device for agricultural pests
CN114451380B (en) * 2022-02-17 2022-12-06 安徽青云峡旅游发展有限公司 Monitoring device and method for scenic spot
CN117099757B (en) * 2023-08-25 2024-04-23 肇庆市承铭农业技术开发有限公司 Method for monitoring solenopsis invicta by using solenopsis invicta prevention device
CN117625381A (en) * 2024-01-26 2024-03-01 深圳和合医学检验实验室 Sampling device and method for virus detection

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102125011A (en) * 2010-01-15 2011-07-20 万德福生物科技股份有限公司 Smell attractant adhesive paper using contrast color
CN102288606A (en) * 2011-05-06 2011-12-21 山东农业大学 Pollen viability measuring method based on machine vision
JP2012065676A (en) * 2012-01-10 2012-04-05 Lintec Corp Insect trap
CN104186449A (en) * 2014-08-15 2014-12-10 北京农业信息技术研究中心 Pest monitoring system capable of automatically replacing pest sticky board and monitoring method
CN204949213U (en) * 2015-09-14 2016-01-13 苏州众人健身器材有限公司 Self -luminous formula ground insecticidal device
CN110287882A (en) * 2019-06-26 2019-09-27 北京林业大学 A kind of big chrysanthemum kind image-recognizing method based on deep learning

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102125011A (en) * 2010-01-15 2011-07-20 万德福生物科技股份有限公司 Smell attractant adhesive paper using contrast color
CN102288606A (en) * 2011-05-06 2011-12-21 山东农业大学 Pollen viability measuring method based on machine vision
JP2012065676A (en) * 2012-01-10 2012-04-05 Lintec Corp Insect trap
CN104186449A (en) * 2014-08-15 2014-12-10 北京农业信息技术研究中心 Pest monitoring system capable of automatically replacing pest sticky board and monitoring method
CN204949213U (en) * 2015-09-14 2016-01-13 苏州众人健身器材有限公司 Self -luminous formula ground insecticidal device
CN110287882A (en) * 2019-06-26 2019-09-27 北京林业大学 A kind of big chrysanthemum kind image-recognizing method based on deep learning

Also Published As

Publication number Publication date
CN110742047A (en) 2020-02-04

Similar Documents

Publication Publication Date Title
CN110742047B (en) Automatic paper changing system based on computer vision effect and insect pest monitoring system
CN104186449B (en) A kind of worm monitoring system and monitoring method of automatic replacing insect-sticking plate
US11659826B2 (en) Detection of arthropods
Wemmer et al. Mammalian sign
CN103168762B (en) Image capture device of field light for luring insects
WO2021250444A1 (en) Insect monitoring system and method
Nazir et al. Advances in image acquisition and processing technologies transforming animal ecological studies
CN103179993A (en) A real-time insect monitoring device
JP3796526B2 (en) Pest counting device
Kawanishi Population status of tigers (Panthera tigris) in a primary rainforest of Peninsular Malaysia
CN111642478B (en) Pest identification system
CN107018361A (en) The method and system of violated fishing gear are recognized based on unmanned plane and graphical analysis
EP3516580A1 (en) Control of harmful organisms
Chiwamba et al. An application of machine learning algorithms in automated identification and capturing of fall armyworm (FAW) moths in the field
CN113475477A (en) Energy-saving air-suction type insect killing lamp trapping device and trapping method
CN104872094A (en) Automatic insect condition observing and predicting equipment
CN204670197U (en) A kind of automation detecting and reporting pest information equipment
AU2021103499A4 (en) An automated plantation health monitoring system and a method thereof
CN213908084U (en) Full-automatic remote monitoring instrument for aedes eggs
Hoogenboezem et al. A model for switching between particulate-feeding and filter-feeding in the common bream, Abramis brama
CN109684960A (en) The determination method and system of the spraying time of defoliant
CN115424196A (en) Tracking monitoring system based on artificial intelligence
CN203207012U (en) Field insect light trap image acquisition device
CN201051837Y (en) Automatic flies trapping device
CN219500247U (en) Remote continuous trapping and killing monitoring instrument for adult aedes

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