KR101572207B1 - Automatic counting device for the Scirtothrips Dorsalis - Google Patents

Abstract

An apparatus for automatically counting convex shed worms is disclosed. The present invention relates to a device for automatically counting Scirtothrips Dorsalis, a micro-insect pest, which has recently been found to be damaging fruit trees, using Machine Vision and image processing techniques. The convex wormworm automatic counting apparatus includes a trap for collecting convex whiskers, a trap stage on which the trap is seated and movable back and forth, left and right, an illuminating unit for illuminating the trap, a camera for transmitting the trap image, A lens which is installed between the camera and the trap so as to be able to raise and lower the focus, and a counting unit which receives the image acquired by the camera and discriminates the convex wormhole and counts the number.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic counting device for the Scirtothrips Dorsalis,

More particularly, the present invention relates to a device for automatically counting convex shed worms, and more particularly, to a method and apparatus for detecting a microscopic pest, Scirtothrips Dorsalis, To a device for automatically counting using technology.

Generally, pests cause many damages to crops. For example, insect pests, such as thrips, can damage fruit and reduce the yield of fruit and decrease the commercial quality of the fruit.

Therefore, in order to prevent damage to crops caused by insect pests, it is important for the administrator to know in advance the poultry populations related to the crops and the pest density around the crops.

However, most pests are very active species, and they are often difficult to see with the naked eye because they are small in size.

Therefore, recently, a pest trapping trap apparatus has been developed in which pheromone and glue are used to catch pests in a attracted state to facilitate the inspection of pests. That is, the conventional insect trapping trap apparatus is composed of a main body forming a manned space and a sticky body attached to the bottom surface of the main body.

When the insect trapping apparatus of this type is used, the pheromone is released through the pheromone emitter on one side of the main body. When the insect trapping apparatus of this state is installed at a certain place of the cultivation plant, the insect pest attracted to the attracting space by the pheromone, And the user can easily grasp the occurrence situation of the individual by observing and counting the trapped pests in the main body.

However, among the tropical insect species of which 60 species are distributed in Korea, the protruding sperm worms, which have recently been infested from Southeast Asia and mainly suffered from the celery root of Jeju Island, have adult sizes of less than 1 mm, Can not be identified with the naked eye.

On the other hand, a micro insect having a size of less than 1 mm, which is limited to observation and counting even by using such a conventional insect trapping apparatus, must observe the entire trap through an optical microscope to identify and count the insect. Such a forecasting method requires a long time for identification and requires a separate manpower to identify a pest.

In agriculture, fruit crops (citrus fruits, apples, pears, tomatoes, etc.) cultivated in houses and ponds are periodically monitored for pests. The inspections on insect pests are mainly carried out by installing simple traps (sticky traps, sex pheromone traps, etc.), inspecting the number of trapped insects directly at the site, collecting traps, or counting them at a designated place.

Among pests, there are microscopic insects (less than 1 mm) small enough to be discernible to the naked eye, such as convex sperm worms. The inspections for microbial pests can not be counted in the field and are counted by separate grafts where they have optical microscopes. This type of forecasting has many limitations in the identification time and manpower.

In addition, there is a difference in identifying the convex sphagnum depending on the skill level of the user, and if the wide trap is enlarged by an optical microscope, some parts may be missed, resulting in a problem of poor accuracy and reliability in the identification of the population .

Korean Patent No. 10-1109337

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and it is an object of the present invention to provide a device for acquiring a pest tracing trap as an image and automatically counting convex worms through image recognition.

As means for solving the problem, the present invention provides a trap for collecting a convex sphagnum; A trap stage on which the trap is seated and movable back and forth and left and right; An illumination unit for illuminating the trap; A camera for transmitting the image of the trap; A lens installed between the camera and the trap so as to be able to raise and lower the focus; And a counting unit that receives the image acquired by the camera and discriminates the convex wormhole and counts the number.

Preferably, the apparatus further includes a trap fixing guide for pressing both ends of the trap to fix the trap to the trap stage.

Preferably, the trap fixing guide is rotatably mounted on the trap stage by a hinge, and the hinge is provided on the torsion spring so that the trap fixing guide fixes the trap in close contact with the trap stage.

Preferably, the trap stage is provided with a lever for rotating the trap fixing guide by pushing the inclined surface provided on the trap fixing guide.

Preferably, the lever is provided on the trap stage so as to be rotatable about a vertical rotation axis, and has a pushing portion for pushing the inclined surface at the one end and a knob at the other end.

According to another aspect of the present invention, there is provided an image processing apparatus comprising: a driving unit having a simple trap and a mechanism required to acquire an image; A control unit for moving the trap stage of the driving unit at a fine interval; An image unit adapted to capture a trap image; An illuminator configured to illuminate the trap; And a counting unit counting the number of traps captured using the image recognition.

Preferably, in the automatic counting device for convex yellow worms according to the present invention, the driving unit attaches a movable guide adjustable according to the size of a trap through a two-axis (X-Y) movable trap stage to improve user friendliness and ease of image acquisition. Particularly, the trap installed in the hearth is equipped with a trap fixing guide to solve the problem that the image is not recognized properly because the trap is irregularly deformed or turned due to the natural environment,

Preferably, in the automatic convex insect worm counting apparatus according to the present invention, the driving unit adjusts the focal distance between the trap and the lens through a vertical (Z-axis) movable motor to correct the quality of the acquired image, And provides an appropriate image.

Preferably, in the convex yellow worm automatic counting apparatus according to the present invention, the driving unit limits a moving range of the trap stage through a limit sensor, thereby preventing rotation overload of the driving motor and providing a trap image acquisition range .

Preferably, in the convex yellow worm automatic counting apparatus according to the present invention, the control unit controls the motor and the sensor mounted on the driving unit through the motion controller and the sensor driver, and performs smooth motion control through communication with the counting unit .

Preferably, in the automatic convex insect worm counting apparatus according to the present invention, the image portion is mounted with a micro-lens through which insects of less than 1 mm can be displayed on a field of view (FOV) So that the image processing using the color values in the image data can be performed.

Preferably, in the convex yellow worm automatic counting apparatus according to the present invention, the illuminating unit is provided with a roughness controller so that the illuminance suitable for the use environment can be digitally controlled within the range of 0 to 1024.

Preferably, in the convex whisker automatic counting apparatus according to the present invention, the counting unit learns the shape of the convex wormhole through the artificial neural network to generate a discrimination model, and performs discrimination by utilizing it in the recognition process , And in particular, a gray image and a gradient image are synthesized in a preprocessing process to improve a recognition rate.

The present invention as described above has the following effects.

The convex sphagnum insect automatic counting apparatus according to the present invention reduces the time required for counting insects, eliminates the case where the user mistakenly identifies them, and is suitable to easily utilize the unskilled users.

1 is a block diagram showing a configuration of an automatic counting apparatus for a convex whitewash according to a preferred embodiment of the present invention.
FIG. 2 is a front view of a convex hermit crab automatic counting apparatus according to a preferred embodiment of the present invention.
FIG. 3 is a front view and a side view of a convex sphagnum automatic counting apparatus according to a preferred embodiment of the present invention.
FIGS. 4 and 5 are perspective views of a trap fixing guide which is a component of the automatic counting device for convex whiskers according to the present invention. FIG. 4 shows the trap fixing position and FIG. 5 shows the trap releasing position.
FIG. 6 is an exemplary view showing a method of dividing an image into traps placed on a trap stage in a convex whitewash automatic counting apparatus according to a preferred embodiment of the present invention.
FIG. 7 is a flowchart illustrating a procedure for acquiring a trap image in a convex worm-worm automatic counting apparatus according to a preferred embodiment of the present invention.
FIG. 8 is a flowchart showing a procedure for automatically counting the automatic counting device of convex amphibious worms according to a preferred embodiment of the present invention.
FIG. 9 is a flowchart showing a procedure for generating a discrimination model in the convex whitewash automatic counting apparatus according to a preferred embodiment of the present invention.
FIG. 10 is a flowchart illustrating a procedure for discriminating an image in a convex yellow worm automatic counting apparatus according to a preferred embodiment of the present invention.
FIG. 11 is a flowchart showing a procedure of a preprocessing process in a convex whitbill automatic counting apparatus according to a preferred embodiment of the present invention.

The above objects, features and advantages of the present invention will become more apparent from the following detailed description. Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

A convex sphagnum automatic counting apparatus according to a preferred embodiment of the present invention is shown in Figs.

1, a convex yellow worm automatic counting apparatus according to the present invention includes a driving unit 100, a control unit 200, an image unit 300, and an illumination unit 400, and includes a counting unit 500, . ≪ / RTI >

As shown in FIG. 2, the driving unit 100 includes a trap stage 110 on which traps on which insects are trapped, a driving motor 110 for providing a driving force for moving the stage in the X- and Y- And a motor booth 120 incorporating a limit sensor capable of limiting the range of movement of the stage 110. [

Here, the trap stage 110 includes a trap fixing guide 111 for fixing a trap that is irregularly bent or twisted.

The trap fixing guide 111 is shown in Figs. 4 and 5. FIG. 4 shows a state in which the trap 600 is unfolded and fixed, and FIG. 5 shows a state in which the trap 600 is released.

The trap fixing guide 111 is provided with two pieces so that the trap 600 can be unfolded from both sides.

The trap fixing guide 111 is rotatably installed on the trap stage 110 by a hinge 113 and a torsion spring 114 for providing an elastic force for pressing and holding the trap 600 is mounted on the hinge 113 Respectively.

The trap fixing guide 111 is provided with a lever 112 for each trap fixing guide 111 so as to be manually operated.

The lever 112 is rotatably mounted on a rotation axis 112c perpendicular to the trap stage 110 and has a handle 112a at one end thereof. And a pushing portion 112b for pushing the inclined surface 111a of the trap fixing guide 111 to rotate the trap fixing guide 111 is provided at the other end of the lever 112. [

4, the trap fixing guide 111 unfolds the trap 600 by the elastic force of the spring 114, and the lever 112 is engaged with the inclined surface 111a of the trap fixing guide 111 It is not pressed. In the case of FIG. 5, the operator turns the handle 112a of the lever 112 clockwise to set the trap fixing guide 111 upright. When the operator turns the lever 112 as described above, the pushing portion 112b of the lever 112 is pushed up along the inclined surface 111a of the fixed guide 111, Is in a state as shown in FIG. 5, and it is rotated to overcome the elastic force of the spring 114 and stands vertically.

As a result, the trap 600 is released from the trap fixing guide 111 and can be replaced with a new trap.

The driving unit 100 includes a support base 130 for mounting the image unit 300 and a support base 130 for fixing the support base 141 on which the illumination unit 400 can be mounted. And a switch booth 140 that collects switches for performing blocking and emergency stop operations.

And a main switch 141 for supplying or cutting off the entire power of the driving unit 100 and includes a monitor support 150 and a monitor 151 for confirming the operation of the counting unit 500. Here, the image bearing stand 131 includes a vertically moving focus adjusting motor 121 for adjusting the focal distance between the trap and the lens 320. [

The control box 210 can be used to load the components of the drive unit 100 and the control unit 200 and a roller 160 to form the outer shape of the driving unit 100.

The controller 200 includes a programmable motion controller mounted inside the control box 210 to control the two-axis (XY) movement of the trap stage 110 and to control the vertical movement of the focus adjusting motor 121, A sensor driver is provided which converts a signal from an installed limit sensor into a digital signal in order to limit the range of each motor, And a communication port for receiving a position control command and a speed control command of the counting unit 500 are provided.

Here, the communication port is limited to a serial communication such as USB or RS232C.

The image unit 300 includes a high-resolution color CCD 310 that supports at least 2 megapixels suitable for identifying characteristics of pests less than 1 mm in traps trapped by insects, And a microlens 320 which can correspond to an enlarging effect of magnification.

Here, the image acquired through the image unit 300 is transferred to the counting unit 500 and used in the image providing step S810 of the flowcharts shown in FIGS.

Particularly, in order to more accurately count the number of individuals in the counting unit 500, since the focus of the acquired image needs to be clear, the focus adjusting motor 121 attached to the image capturing unit 131 is pre- It focuses on driving.

The illumination unit 400 may include a light source provided by an LED or the like and a diffusion plate 410 configured to uniformly disperse light. The illumination unit 410 should not be irradiated to the entire trap, , But it is provided as a reflection light so that no shadow is generated when the insect is irradiated. In particular, the illumination controller includes a luminance controller that can adjust the brightness of the illumination according to the illumination control command of the counting unit 500.

The counting unit 500 counts the number of insects according to the image acquisition instruction of the counting unit 500, as shown in the flowcharts of FIGS. 7 to 11.

Referring to FIG. 6, an example of a method of counting the trap 600 will be described. The trap 600 is divided into a plurality of cells 620, and the CCD 310 may be moved along the arrow direction 630 and counted.

FIGS. 7 to 11 illustrate a procedure for performing an image processing operation using the convex whisker automatic counting apparatus according to the present invention.

Referring to FIG. 7, a flow chart of the movement of the trap stage 110 is shown. The trap stage 110 is first moved to the initial position (S710). The next image is acquired (S720), the position is corrected (S730), and finally the trap 610 is moved to the replacement position (S740) so that the trap 610 can be replaced.

8 is a flow chart for image processing. When the image is provided (S810), the image is processed (S820), and the image processing will be described later. The image processing result image is discriminated (S830), and if it is convex sham worm, it is accumulated in the increase of the number of the population (S840), or otherwise, the image is repeated again.

9 shows a learning process for image processing. As shown, the image is received (S910) and preprocessed (S920). Then, the processed image is classified into a correct answer (S930), and the feature is extracted (S940). The discrimination model is generated and stored according to the extracted information (S950).

Fig. 10 shows an image discrimination based on the learning result. If an image is provided (S810), preprocessing is performed (S960), and features are extracted (940). Then, comparison with the above-described discrimination model and discrimination (S970) are ended.

11 is a flowchart showing the above-described preprocessing process. As shown in the figure, when the image is provided (S810), the preprocessing process generates a gray image (S921) and generates a gradient image again (S922). Thereafter, the image is synthesized (S923), the histogram is calculated (S924), and the candidate image is output (S925) to terminate the preprocessing. This preprocessing enables feature extraction with reference.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the inventions. It will be apparent to those of ordinary skill in the art.

100:
110: trap stage 111: trap fixing guide
112: lever 200:
300: video part 400: illuminating part

Claims (5)

Traps collected with convex whiskers;
A trap stage on which the trap is seated and movable back and forth and left and right;
An illumination unit for illuminating the trap;
A camera for acquiring and transmitting an image of the trap;
A lens installed between the camera and the trap so as to be able to raise and lower the focus;
A counting unit that receives the image acquired by the camera and discriminates the convex wormhole and counts the number;
A convex sphagnum automatic counting device. The method according to claim 1,
And a trap fixing guide for pressing both ends of the trap to fix the trap to the trap stage. 3. The method of claim 2,
Wherein the trap fixing guide is rotatably installed on the trap stage by a hinge and the hinge is provided on the torsion spring so that the trap fixing guide fixes the trap tightly to the trap stage. Device. The method of claim 3,
Wherein the trap stage is provided with a lever for rotating the trap fixing guide by pushing the inclined surface provided in the trap fixing guide. 5. The method of claim 4,
Wherein the lever is rotatably mounted on the trap stage about a vertical rotation axis and has a handle at one end and a pushing portion at the other end to push the slope.

Images