CN110689507A

CN110689507A - Insect type identification device, system and method

Info

Publication number: CN110689507A
Application number: CN201911088489.1A
Authority: CN
Inventors: 郭银波; 柳涛; 荚庆
Original assignee: Beijing Softcom Smart City Technology Co Ltd
Current assignee: Beijing Softcom Smart City Technology Co Ltd
Priority date: 2019-11-08
Filing date: 2019-11-08
Publication date: 2020-01-14

Abstract

The embodiment of the invention discloses insect type identification equipment, system and method. The equipment comprises an artificial intelligence AI computing unit, a camera module, a communication module and a pest control device; the camera module is used for shooting an image containing insects; the AI computing unit is used for carrying out edge detection and positioning on insects in the image to obtain an insect image, segmenting a background image and the insect image in the image, segmenting at least two types of insect images, and extracting morphological characteristics and color characteristics of the insects from the segmented image; comparing the insect characteristics with prestored insect characteristics, and sending a comparison result to a cloud server through a communication module so that the cloud server sends an insect disease alarm signal based on the comparison result to warn workers; and the pest control device is used for controlling pests according to the comparison result. This equipment can in time discern the insect species, conveniently masters the crops condition, can prevent and treat the plant diseases and insect pests, uses manpower sparingly.

Description

Insect type identification device, system and method

Technical Field

The embodiment of the invention relates to a crop pest type identification and control equipment technology, in particular to insect type identification equipment, system and method.

Background

The diseases and insect pests of crops are one of the main agricultural disasters, and the agricultural production and the economic development are seriously influenced. If the insect species in the crops are effectively identified, the method is beneficial to adopting corresponding prevention measures for different plant diseases and insect pests, can prevent diseases and insect pests from growing slightly, and can effectively inhibit the spread and propagation of the plant diseases and insect pests to a certain extent. Therefore, the method is very important for controlling the plant diseases and insect pests of the crops by collecting the insect images of the crops and identifying the insect species in time and preventing and controlling the insect species.

At present, after insect investigation of crops and judgment of insect species by naked eyes are generally carried out, corresponding measures are taken to inhibit the spread and spread of plant diseases and insect pests. The mode is time-consuming and labor-consuming, has large subjective factor influence, and is not beneficial to timely mastering the pest and disease kinds of crops. And after insect species identification is carried out by collecting insect images in a lamp-luring mode, corresponding measures are taken to inhibit the spread and propagation of plant diseases and insect pests. The mode can attract various and most insects, even some diseases and insect pests which do harm to other crops but do not harm the current crops, can not reflect the real disease and insect pest conditions of the current crops, and is not beneficial to insect species identification. In addition, the two modes require artificial pest control measures, consume manpower, and have certain influence on human health when the pest control measures are manually taken.

Disclosure of Invention

The invention provides an insect type identification device, system and method, which can identify insect types in time, facilitate the mastering of crop conditions, prevent and treat plant diseases and insect pests and save manpower.

In a first aspect, an embodiment of the present invention provides an insect type identification device, including an Artificial Intelligence (AI) calculation unit, a camera module, a communication module, and a pest control apparatus;

the camera module is used for shooting an image containing insects;

the AI computing unit is electrically connected with the camera module and is used for carrying out edge detection on insects in the image, extracting texture features to position the insects in the image to obtain an insect image, segmenting a background image and the insect image in the image and at least two types of insect images, and extracting morphological features and color features of the insects from the segmented image; comparing the extracted features with prestored insect features, and sending a comparison result to a cloud server through the communication module, so that the cloud server sends a pest and disease alarm signal based on the comparison result to warn workers; wherein the morphological features include shape features and transform domain features; the shape characteristics include area, perimeter, main axis direction, tightness, eccentricity and curvature;

and the pest control device is used for controlling pests according to the comparison result.

Optionally, the apparatus further comprises: a position location module;

the position locating module is used for identifying the distance between the insect and the pest control device and the direction of the insect relative to the pest control device.

Optionally, the pest control device comprises at least one of a drug spray device, a radiation generation device, or a light, smell, and color control device.

Optionally, the apparatus further comprises: a light supplement device;

and the light supplementing device is used for supplementing light to the camera module when the brightness of the external environment light is less than the set brightness.

Optionally, the communication module includes: the system comprises a narrow-band Internet of things NB-IOT module and an antenna ANT;

the NB-IOT module is electrically connected with the AI computing unit and used for uploading the comparison result to a cloud server through the ANT.

Optionally, the apparatus further comprises: the power management system comprises a memory module, a battery unit and a power management chip PMIC;

the storage module is electrically connected with the camera module and the AI computing unit respectively; the image acquisition module is used for receiving and storing the image sent by the camera module;

the battery unit is electrically connected with the PMIC and used for providing electric energy for the PMIC;

the PMIC is electrically connected with the AI computing unit, the communication module, the storage module and the pest control device respectively and is used for managing electric energy provided by the battery unit to the AI computing unit, the communication module, the storage module and the pest control device.

Optionally, the apparatus further comprises: an insect detection module;

the insect detection module is used for starting the camera module to shoot an image containing the insects when the insects are detected to appear; and when the disappearance of the insects is detected, closing the camera module.

Optionally, the apparatus further comprises: a timing module;

the timing module is used for indicating the pest control device to stop pest control when first preset time is reached; and instructing the insect monitoring module to start detecting when a second preset time is reached.

In a second aspect, an embodiment of the present invention further provides an insect type identification system, where the system includes the insect type identification device according to any embodiment of the present invention, and further includes a cloud server;

and the cloud server is used for receiving the comparison result sent by the communication module and sending a pest and disease damage alarm signal based on the comparison result.

In a third aspect, an embodiment of the present invention further provides an insect type identification method, which is applied to the insect type identification apparatus according to any embodiment of the present invention, and includes:

shooting an image containing the insect through a camera module;

carrying out edge detection on insects in the image through an artificial intelligence AI computing unit, extracting texture features to position the insects in the image to obtain an insect image, segmenting a background image and the insect image in the image and segmenting at least two types of insect images, and extracting morphological features and color features of the insects from the segmented image; comparing the extracted features with prestored insect features, and sending a comparison result to a cloud server through the communication module, so that the cloud server sends a pest and disease alarm signal based on the comparison result to warn workers; wherein the morphological features include shape features and transform domain features; the shape characteristics include area, perimeter, main axis direction, tightness, eccentricity and curvature;

and performing pest control according to the comparison result by using a pest control device.

The technical scheme provided by the embodiment of the invention provides insect type identification equipment, which comprises an AI computing unit, a camera module, a communication module and a pest control device; shooting an image containing the insect through a camera module; the method comprises the steps that an AI computing unit is used for carrying out edge detection on insects in an image, realizing positioning, background and insect segmentation and segmentation among different insects, extracting morphological characteristics and color characteristics of the insects to be compared with prestored insect characteristics, and sending a comparison result to a cloud server through a communication module so that the cloud server sends an insect disease and pest alarm signal based on the comparison result; and pest control is carried out on the basis of the alarm signal according to the cloud server through the pest control device. The problem of the kind discernment of insect in the crops and the prevention and cure of plant diseases and insect pests in the crops is solved, realized under multiple scene, in time discerning the insect kind, the convenient crop condition of in time mastering can prevent and cure the plant diseases and insect pests, need not artificial operation, uses manpower sparingly, does benefit to the effect of health.

Drawings

Fig. 1 is a schematic structural diagram of an insect type identification device according to an embodiment of the present invention;

FIG. 2 is a schematic structural diagram of an insect type identification device according to a second embodiment of the present invention

Fig. 3 is a schematic structural diagram of an insect type identification system according to a third embodiment of the present invention;

fig. 4 is a flowchart of an insect type identification method according to a fourth embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.

Example one

Fig. 1 is a schematic structural diagram of an insect type identification device according to an embodiment of the present invention, which is applicable to the case of identifying the type of insects in crops and/or preventing and controlling plant diseases and insect pests of crops. As shown in fig. 1, the insect type identifying apparatus 100 includes: an AI calculation unit 121, an image pickup module 111, a communication module 123, and a pest control device 130.

The camera module 111 is used for shooting an image containing insects;

alternatively, the camera module 111 may be disposed on the sensor board 110, the camera module 111 may be a Complementary Metal Oxide Semiconductor (CMOS) camera, and the pixels may be 30 ten thousand pixels, and may capture an image containing insects with a resolution of 640 × 480. The camera module 111 may be connected to an image encoding chip (not shown in fig. 1), and may perform encoding processing on a captured image, for example, a picture may be processed into a JPG format.

The AI calculation unit 121 is electrically connected to the camera module 111, and is configured to perform edge detection on insects in the image, extract texture features to locate the insects in the image, obtain an insect image, segment a background image in the image and the insect image, segment at least two types of insect images, and extract morphological features and color features of the insects from the segmented image; comparing the extracted features with prestored insect features, and sending a comparison result to a cloud server through a communication module 123 so that the cloud server sends a pest alarm signal based on the comparison result; the morphological features include shape features and transform domain features; wherein the shape characteristics include area, perimeter, major axis direction, tightness, eccentricity and curvature;

among them, the AI calculation unit 121 may be provided on the main control board 120. The edge is the most basic feature of the image, is a pixel set with the gray level of the surrounding pixels having step change or roof change, and is an important basis for image segmentation, region shape feature extraction and target region identification. Edge detection can be performed by using a Sobel operator, a Roberts operator, a Canny operator, a Log operator, a Prewitt operator, a Laplace operator, or the like. The energy, entropy, moment of inertia and local stable 4 parameter values of the texture features can be extracted through different operators, and edge detection of the insects such as butterflies and aphids is achieved. The edge detection textural property of the Laplace operator is superior to that of other operators. The localization of the insects in the image can be achieved by edge detection.

The image segmentation is a process of dividing an image into mutually disjoint regions according to certain characteristics of the image and extracting a key interested part. Therefore, whether the information interested in the insect image segmentation can be effectively obtained directly influences the accuracy of the final identification and counting result. The image segmentation includes segmentation between a background image and an insect image and segmentation between various insects. The segmentation method between the background image and the insect image can adopt a threshold value method, an edge flow method, a wavelet analysis method and the like. The image segmentation of the grain storage insects and the lepidoptera insects can be realized by adopting methods such as a histogram threshold method, a self-adaptive threshold method, a relative entropy threshold method and the like.

The image feature extraction is a process of converting visual features of an image into a mathematical form which can be identified by a computer, and provides a reliable basis for automatic identification and counting of target insects. The extracted insect image features generally include color features, morphological features, and the like. The image color features are the most widely applied visual features, the dependency on the size, direction and visual angle of the image is small, and therefore the butterfly color pattern features have high robustness, and the butterfly color pattern features are obtained by using 4 pieces of one-dimensional color histogram information of Red (Red, R), Green (Green, G), Blue (Blue, B) and brightness (Light, L) and R, G pieces of two pieces of chromaticity two-dimensional histogram information which are independent of illumination, front and back color feature values are respectively extracted, and the butterfly color pattern features can be automatically identified by combining with a neural network. However, the color statistics of different insects are likely to be similar, and the shapes formed by scales of various colors on the wings of many lepidoptera insects cannot be identified only by counting the probability of each color pixel. Therefore, on the basis of extracting the color pixels, the color feature vectors of the spatial positions can be combined, and the lepidoptera insects can be identified.

Among them, morphological features are used more in insect identification and can be divided into shape features and transform domain features. The shape characteristic parameters can be area, perimeter, main axis direction, tightness, eccentricity, curvature and the like. The morphological characteristics are suitable for identifying insects with complete morphologies and insects with large individual and morphological differences, but the completeness of an insect sample and the sample placement consistency seriously affect the automatic insect identification result based on the morphological characteristics, and meanwhile, the morphological characteristics of live insects in different states show differences, so that the identification of insects by only utilizing the morphological characteristics is limited to a certain extent.

Different insects have different characteristics such as color, shape, texture and the like, and the characteristic extraction of a single bottom layer ignores the connection among multiple characteristics and the full understanding of various forms of information of the image. The AI calculation unit 121 may extract morphological features such as rectangularity and HU invariant moments of insects such as cicadas and chafer aeruginosa, and color features such as a gray level histogram and a two-dimensional chromaticity histogram, and implement remote insect identification based on a Browser/Server (B/S) structure by combining a radial basis neural network classifier and an Internet network. The identification rates of 3 kinds of single characteristics of color, form and texture for identifying the insects such as the butterfly, the spodoptera and the like are respectively 75%, 78.2% and 82%, and the identification rate based on the comprehensive characteristics of the color, the form and the texture can be 90%. A combined model based on global features and local features is established, and classification and identification of insects are effectively achieved. During the discernment can compare the characteristic of drawing with the insect characteristic of prestoring to with the comparative result, also be exactly the identification result, send the high in the clouds server through communication module 123, so that the high in the clouds server sends pest alarm signal based on the comparative result.

Optionally, as shown in fig. 1, the communication module 123 includes: a narrowband Band internet of Things (NB-IOT) module and Antenna (ANTENNA, ANT) (neither shown in FIG. 1);

the NB-IOT module is electrically connected to the AI calculation unit 121, and is configured to upload the comparison result to a cloud server (not shown in fig. 1) via an ANT.

It should be noted that the communication module 123 may be disposed on the main control board 120, and the NB-IOT module and the ANT may be used to make the insect type identifying device 100 have low power consumption and low cost, and may prolong the service life of the insect type identifying device 100. The cloud server can send a pest alarm signal when the comparison result of the insects is a certain type of pest, so that a crop manager can be reminded of checking the condition of crops in time. The alarm signal may be a buzzer sounding, an indicator light lighting, or a voice prompt comparison result, and the invention is not particularly limited.

And a pest control device 130 for performing pest control according to the comparison result.

Wherein, optionally, the pest control device 130 comprises at least one of a drug spray device, a radiation generating device, or a light odor color control device.

It should be noted that the pest control device 130 may be electrically connected to the communication module 123, and may take a certain pest control measure for different types of pests based on the alarm signal according to the comparison result or according to the cloud server. The pesticide spraying device can eliminate most of plant diseases and insect pests by spraying pesticides; the radiation generating device can make some plant diseases and insect pests lose fertility by generating radiation and can not reproduce offspring, so that the aim of killing the plant diseases and insect pests is achieved; the lamplight smell and color control device can trap and kill plant diseases and insect pests through a combined trapping and killing method of light attraction, color attraction and flavor attraction, and can also be assisted by combining with the arrangement of an insect-proof net. These pest control measures are harmful to human health, and the adoption of the insect type identifying device 100 can save manpower and is beneficial to human health by means of the device without adopting the pest control measures manually.

The technical scheme of the embodiment of the invention provides insect type identification equipment, which comprises an AI computing unit, a camera module, a communication module and a pest control device; shooting an image containing the insect through a camera module; the method comprises the steps that an AI computing unit is used for carrying out edge detection on insects in an image, realizing positioning, background and insect segmentation and segmentation among different insects, extracting morphological characteristics and color characteristics of the insects to be compared with prestored insect characteristics, and sending a comparison result to a cloud server through a communication module so that the cloud server sends an insect disease and pest alarm signal based on the comparison result; and pest control is carried out by the pest control device according to the comparison result or according to the cloud server based on the alarm signal. The problem of the kind discernment of insect in the crops and the prevention and cure of plant diseases and insect pests in the crops is solved, realized under multiple scene, in time discerning the insect kind to can produce alarm signal according to the comparative result, conveniently in time master the crops condition, can prevent and cure the plant diseases and insect pests, need not artificial operation, use manpower sparingly, do benefit to the effect of health.

On the basis of the above embodiment, optionally, as shown in fig. 1, the insect type identifying apparatus 100 further includes: an insect detection module 114;

an insect detection module 114, configured to activate a camera module to capture an image containing an insect when the presence of the insect is detected; and when the insect disappears, the camera module is closed.

Wherein, insect detection module 114 can set up on sensor board 110, and insect detection module 114 can detect whether the insect appears through sensor class equipment, radar monitoring method or soft X-ray machine perspective detection method. For example, local temperature elevation in the crop can be detected by a temperature sensor, and the occurrence of the insect is determined; alternatively, the presence of insects can be detected by radar signals reflected by the insects; alternatively, the presence of insects can be detected by an agricultural soft X-ray machine; when the insect is detected, the camera module 111 can be instructed to be started to shoot the image containing the insect; when the disappearance of the insects is detected, the camera module 111 may be instructed to turn off. The camera module 111 can be prevented from being in a working state all the time, the camera module 111 is protected, the service life of the camera module 111 is prolonged, and therefore the service life of the insect type identification device 100 is prolonged.

On the basis of the above embodiment, optionally, as shown in fig. 1, the insect type identifying apparatus 100 further includes: a light supplement device 112;

and a light supplement device 112, configured to supplement light to the camera module 111 when the brightness of the external environment light is less than the set brightness.

The light supplement device 112 may be disposed on the sensor board 110, may be a Light Emitting Diode (LED) light supplement lamp, and may be turned on when the brightness is low, and the auxiliary camera module 111 captures a clearer image containing the insects; and when the brightness is high, the switch is switched off. The problem that the processing of later-stage images and the identification difficulty of insect species are caused by poor effect of the shot images containing insects when the brightness is low can be avoided. The light supplement device 112 can be turned on or off, so that resources can be saved.

On the basis of the above embodiment, optionally, as shown in fig. 1, the insect type identifying apparatus 100 further includes: a position-location module 113;

a position locating module 113 for identifying a distance of the insect from the pest control device 130 and a direction of the insect relative to the pest control device 130.

Wherein, position location module 113 can set up on sensor board 110, and disease and pest control device 130 can realize accurate prevention and cure to the disease and pest in the crops according to the distance of insect with the disease and pest control device and the direction of insect for the disease and pest control device. By way of example, the pest control device 130 can use the distance and direction identified by the position locating module 113 as the center of a circle to perform pest control on the area with a certain radius, so that the pest control device 130 is not affected when other crops have no pest, and accurate control can be realized on crops with pests.

On the basis of the above embodiment, optionally, as shown in fig. 1, the insect type identifying apparatus 100 further includes: a timing module 124;

a timing module 124 for instructing the pest control device 130 to stop pest control when a first preset time is reached; and instruct the insect monitoring module 114 to start detecting when the second preset time is reached.

Wherein, timing module 124 can set up on main control board 120, can realize through timing module 124 that pest control device 130 stops when working a period, first preset time is arrived promptly, can avoid excessively preventing and curing, when killing the pest, has also caused harm to crops. And when a period of time elapses again, that is, the second preset time is reached, the insect detection module 114 can be instructed to detect the crops again, and when the insects are detected, the camera module 111 can be instructed to turn on again to shoot the images containing the insects, so that the insects are identified again. The prevention and control of the plant diseases and insect pests can be realized only in a certain time, so that the situation that the insect detection module 114 detects the plant diseases and insect pests all the time in the period can be avoided, the insect is identified again, the plant diseases and insect pests are continuously prevented and controlled, excessive prevention and control are carried out, and the influence on crops is generated. And after the prevention and control of the plant diseases and insect pests play a role, whether the plant diseases and insect pests are completely killed or not can be detected, and if not, the plant diseases and insect pests can be detected, shot, identified and prevented again, so that the plant diseases and insect pests can be completely killed. In addition, the timing module 124 can also prevent the insect detection module 114 from being in an operating state all the time, thereby saving resources and prolonging the service life of the insect type identification device 100.

In one implementation of the embodiment of the present invention, optionally, as shown in fig. 1, the insect type identifying apparatus 100 further includes: a memory module 122, a battery unit 126, and a PMIC 125;

the storage module 122 is electrically connected to the camera module 111 and the AI calculation unit 121 respectively; used for receiving and storing the image sent by the camera module 111; the battery unit 126 is electrically connected to the PMIC125 for providing electric power to the PMIC 125; the PMIC125 is electrically connected to the AI calculation unit 121, the communication module 123, the storage module 122, and the pest control device 130, respectively, and is configured to manage electric energy supplied from the battery unit 126 to the AI calculation unit 121, the communication module 123, the storage module 122, and the pest control device 130.

It should be noted that the storage module 122 may be disposed on the main control board 120, and may be a memory flash using a Serial Peripheral Interface (SPI). Accordingly, the connection between the storage module 122 and the camera module 111 may be an Asynchronous Receiver/Transmitter (UART). The image may be transferred from the camera module 111 to the storage module 122 on the main control board 120 for storage. A battery unit 126 may be provided on the main control board 120 for supplying power to the insect type recognition apparatus 100 so that the apparatus can operate normally. The PMIC125 may be disposed on the main control board 120, and may manage power supplies of devices on the device, so as to ensure that the devices operate under appropriate voltage and current.

Example two

Fig. 2 is a schematic structural diagram of an insect type identification device according to a second embodiment of the present invention, and as shown in fig. 2, the insect type identification device 100 includes: an AI calculation unit 121, an image capture module 111, and a communication module 123.

The camera module 111 is used for shooting an image containing insects;

alternatively, the camera module 111 may be disposed on the sensor board 110, the camera module 111 may be a CMOS camera, the pixels may be 30 ten thousand pixels, and the image including the insect with the resolution of 640 × 480 may be captured. The camera module 111 may be connected to an image encoding chip (not shown in fig. 2), and may perform encoding processing on the captured image, for example, may process a picture into a JPG format.

a combined model based on global features and local features is established, and classification and identification of insects are effectively achieved. During the discernment can compare the characteristic of drawing with the insect characteristic of prestoring to with the comparative result, also be exactly the identification result, send the high in the clouds server through communication module 123, so that the high in the clouds server sends pest alarm signal based on the comparative result.

Optionally, as shown in fig. 2, the communication module 123 includes: NB-IOT module and ANT (neither shown in FIG. 2);

the NB-IOT module is electrically connected to the AI calculation unit 121, and is configured to upload the comparison result to a cloud server (not shown in fig. 2) via an ANT.

The technical scheme of the embodiment of the invention provides insect type identification equipment, which comprises an AI computing unit, a camera module and a communication module; shooting an image containing the insect through a camera module; the AI computing unit is used for carrying out edge detection on insects in the image, realizing positioning, background and insect segmentation and segmentation among different insects, extracting morphological characteristics and color characteristics of the insects and comparing the morphological characteristics and the color characteristics with prestored insect characteristics, and sending comparison results to the cloud server through the communication module so that the cloud server sends pest and disease alarm signals based on the comparison results. The problem of the kind discernment of insect in the crops is solved, realized under multiple scene, in time discerning the insect kind to can produce alarm signal according to the comparative result, conveniently in time master the crops condition, can prevent and treat the plant diseases and insect pests, need not artificial operation, use manpower sparingly, do benefit to the effect of health.

On the basis of the above embodiment, optionally, as shown in fig. 2, the insect type identifying apparatus 100 further includes: a light supplement device 112;

The light supplement device 112 may be disposed on the sensor board 110, may be an LED light supplement lamp, and may be turned on when the brightness is low, and the auxiliary camera module 111 captures a clearer image containing an insect; and when the brightness is high, the switch is switched off. The problem that the processing of later-stage images and the identification difficulty of insect species are caused by poor effect of the shot images containing insects when the brightness is low can be avoided. The light supplement device 112 can be turned on or off, so that resources can be saved.

In one implementation of the embodiment of the present invention, optionally, as shown in fig. 2, the insect type identifying apparatus 100 further includes: a memory module 122, a battery unit 126, and a PMIC 125;

It should be noted that the storage module 122 may be disposed on the main control board 120, and may be a memory flash using an SPI. Accordingly, the connection between the storage module 122 and the camera module 111 may be a UART. The image may be transferred from the camera module 111 to the storage module 122 on the main control board 120 for storage. A battery unit 126 may be provided on the main control board 120 for supplying power to the insect type recognition apparatus 100 so that the apparatus can operate normally. The PMIC125 may be disposed on the main control board 120, and may manage power supplies of devices on the device, so as to ensure that the devices operate under appropriate voltage and current.

EXAMPLE III

Fig. 3 is a schematic structural diagram of an insect type identification system according to a third embodiment of the present invention, and as shown in fig. 3, an insect type identification system 200 includes an insect type identification device 100 according to any embodiment of the present invention, and a cloud server 210.

The cloud server 210 is configured to receive the comparison result sent by the communication module 123, and send a pest alarm signal based on the comparison result to warn a worker.

It should be noted that the cloud server 210 may send a pest alarm signal when the comparison result of the insect is a certain type of pest, so as to conveniently remind the worker to check the condition of the crop in time. The alarm signal may be a buzzer sounding, an indicator light lighting, or a voice prompt comparison result, and the invention is not particularly limited.

The insect type identification system provided by the embodiment of the invention has the corresponding beneficial effects of the insect type identification device provided by any embodiment of the invention.

Example four

Fig. 4 is a flowchart of an insect type identification method according to a fourth embodiment of the present invention, which can be applied to the insect type identification device or the insect type identification system according to any embodiment of the present invention. As shown in fig. 4, the method of the embodiment of the present invention specifically includes:

and S310, shooting an image containing the insect through a camera module.

The camera module can be disposed on the sensor board, can be connected to an image coding chip, and can perform coding processing on a shot image, for example, a picture can be processed into a JPG format.

Optionally, by the insect detection module, when the presence of an insect is detected, the camera module is started to shoot an image containing the insect; and when the insect disappears, the camera module is closed.

Optionally, the light supplement device is used for supplementing light to the camera module when the brightness of the external ambient light is less than the set brightness.

S320, detecting the edges of the insects in the image through an artificial intelligence AI computing unit, extracting texture features to position the insects in the image to obtain an insect image, segmenting a background image and the insect image in the image and at least two types of insect images, and extracting morphological features and color features of the insects from the segmented image; comparing the extracted features with prestored insect features, and sending a comparison result to a cloud server through a communication module so that the cloud server sends a pest alarm signal based on the comparison result; the morphological features include shape features and transform domain features; wherein the shape characteristics include area, perimeter, principal axis direction, tightness, eccentricity, and curvature.

The AI calculation unit may be disposed on the main control board.

Optionally, the communication module includes: an NB-IOT module and an ANT;

and uploading the comparison result to a cloud server through an ANT through an NB-IOT module.

The NB-IOT module is electrically connected with the AI computing unit, and the communication module can be arranged on the main control board.

S330, through the pest control device, pest control is carried out according to the cloud server based on alarm signals.

Optionally, the pest control device comprises at least one of a drug spray device, a radiation generating device, or a light, smell, and color control device.

Optionally, the distance between the insect and the pest control device and the direction of the insect relative to the pest control device are identified by the position locating module.

Wherein the position location module may be disposed on the sensor board.

Optionally, the timing module instructs the pest control device to stop pest control when the first preset time is reached; and instructing the insect monitoring module to start detecting when the second preset time is reached.

Wherein, the timing module can be arranged on the main control board.

Optionally, the image sent by the camera module is received and stored by the storage module;

the storage module is electrically connected with the camera module and the AI computing unit respectively.

Optionally, the PMIC is powered by a battery unit;

wherein the battery unit is electrically connected with the PMIC; the PMIC is electrically connected with the AI computing unit, the communication module, the storage module and the pest control device respectively;

and managing the electric energy provided by the battery unit to the AI computing unit, the communication module, the storage module and the pest control device through the PMIC.

Wherein, the storage module can be arranged on the main control board and can be an SPI flash. Correspondingly, the connection between the storage module and the camera module can adopt UART.

The technical scheme of the embodiment of the invention provides an insect type identification method, which comprises the steps of shooting an image containing an insect through a camera module; the method comprises the steps that an AI computing unit is used for carrying out edge detection on insects in an image, realizing positioning, background and insect segmentation and segmentation among different insects, extracting morphological characteristics and color characteristics of the insects to be compared with prestored insect characteristics, and sending a comparison result to a cloud server through a communication module so that the cloud server sends an insect disease and pest alarm signal based on the comparison result; and pest control is carried out on the basis of the alarm signal according to the cloud server through the pest control device. The problem of the kind discernment of insect in the crops and the prevention and cure of plant diseases and insect pests in the crops is solved, realized under multiple scene, in time discerning the insect kind to can produce alarm signal according to the comparative result, conveniently in time master the crops condition, can prevent and cure the plant diseases and insect pests, need not artificial operation, use manpower sparingly, do benefit to the effect of health.

It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. An insect type identifying apparatus, comprising: the system comprises an artificial intelligence AI computing unit, a camera module, a communication module and a pest control device;

the camera module is used for shooting an image containing insects;

2. The apparatus of claim 1, further comprising: a position location module;

3. The apparatus of claim 1, wherein the pest control device comprises at least one of a medication spray device, a radiation generating device, or a light odor color control device.

4. The apparatus of claim 1, further comprising: a light supplement device;

5. The device of claim 1, wherein the communication module comprises: the system comprises a narrow-band Internet of things NB-IOT module and an antenna ANT;

6. The apparatus of claim 1, further comprising: the power management system comprises a memory module, a battery unit and a power management chip PMIC;

7. The apparatus of claim 1, further comprising: an insect detection module;

8. The apparatus of claim 7, further comprising: a timing module;

9. An insect type identification system comprising the apparatus of any one of claims 1 to 8, and a cloud server;

10. An insect type identification method, comprising:

shooting an image containing the insect through a camera module;