KR102654966B1 - Cultivation Management System - Google Patents

Abstract

The cultivation management system according to an embodiment of the present disclosure includes a water level measuring unit including N circular markers arranged at equal intervals in a direction perpendicular to the surface of the rice field; A cultivation observation device configured to photograph the water level measurement unit; A server configured to receive the image captured by the cultivation observation device and calculate the water level of the rice field; And a user terminal configured to receive the water level calculated by the server through the cultivation observation device and transmit a control command to the cultivation observation device, wherein when the water level measurement unit is submerged in water, M among the N circular markers When several circular markers are submerged in water and a portion of the M+1 circular marker is submerged, the server derives the center of gravity of the non-submerged portion of the M+1 circular marker, and determines the center of gravity of the non-submerged portion. The y-axis value of the hidden part is derived using the y-axis value, and the water level is calculated using the y-axis value of the hidden part.

Description

Cultivation Management System

This disclosure relates to a cultivation management system and a method of using the same. Specifically, it relates to a system that enables cultivation of crops even with a small amount of time and labor by combining the agricultural and IT fields, and allows water level observation in various environments. .

The content described in this section simply provides background information for the present disclosure and does not constitute prior art.

Stable rice cultivation requires many technologies, including sowing, transplanting, fertilization, and pest control. Among them, the thing farmers spend the most time managing during the cultivation period is water management, which involves appropriately controlling irrigation and drainage during the cultivation period.

To manage water during rice cultivation, water channels are formed in the rice fields for water intake and drainage. In order to irrigate a rice field, the water inlet must be opened and closed periodically. A lot of time and labor must be invested in this process, and it is currently the element that requires the most time and labor in rice cultivation.

For such water management, according to the domestic utility model 20-0436680 Y1, a water level sensor is installed to detect the water level in the farmland to detect the lowest and highest points, and a water gate is opened to control the water level by opening and closing the water gate up and down. I'm doing it. Meanwhile, according to the automatic water intake spigot for water management according to the utility model described above, there are problems in that there is no precise water level control function and it is difficult to check malfunction of the device in local packaging.

In addition, workers who are inexperienced in rice cultivation have little knowledge about tillering conditions and seeding conditions during the growing period of rice. Because of this, there is a problem of missing the appropriate harvest time and causing damage.

In addition, in order to measure the water level of a rice field, the worker must directly measure the water level using a measuring tool, which requires excessive labor input, and each worker has different proficiency with the measuring tool, so even if the same rice field is measured, the measurement results are not consistent. There is a problem with not doing it.

In addition, conventional water level markers for measuring the water level of rice fields are generally formed as rulers. When water droplets form on these ruler-type water level markers, the scale becomes distorted, making accurate measurements impossible. Additionally, there is a problem that it is difficult to check the water level marker with the naked eye or with a captured image in a situation where there is no light.

(Utility model document) KR 20-0436680 Y1

(Patent Document) KR 10-1224384 B1

(Patent Document) US 2021/0131852 A1

Accordingly, the purpose of the present disclosure is to provide a cultivation management system capable of controlling hydrology automatically or remotely and manually.

In addition, the present disclosure aims to provide a cultivation management system that can observe the water level of rice paddies in real time and remotely control the water level by automatically or remotely manually opening and closing the valve.

Additionally, the present disclosure aims to provide a cultivation management system that can check device malfunctions in real time.

In addition, the purpose of the present disclosure is to provide a cultivation management system that learns the tillering status and seeding status of rice using an artificial intelligence model and can determine the tillering status and seeding status only by taking crop images.

In addition, the purpose of the present disclosure is to provide a cultivation management system capable of accurately measuring water level during the day, at night, in situations where water is reflected, and in situations where water is condensed.

The problems to be solved by the present invention are not limited to the problems mentioned above, and other problems not mentioned can be clearly understood by those skilled in the art from the description below.

The cultivation management system according to an embodiment of the present disclosure includes a water level measuring unit including N circular markers arranged at equal intervals in a direction perpendicular to the surface of the rice field; A cultivation observation device configured to photograph the water level measurement unit; A server configured to receive the image captured by the cultivation observation device and calculate the water level of the rice field; And a user terminal configured to receive the water level calculated by the server through the cultivation observation device and transmit a control command to the cultivation observation device, wherein when the water level measurement unit is submerged in water, M among the N circular markers When several circular markers are submerged in water and a portion of the M+1 circular marker is submerged, the server derives the center of gravity of the non-submerged portion of the M+1 circular marker, and determines the center of gravity of the non-submerged portion. The y-axis value of the hidden part is derived using the y-axis value, and the water level is calculated using the y-axis value of the hidden part.

In addition, preferably, the cultivation observation device of the present disclosure includes a camera configured to photograph crops grown in the rice field, irrigation water pipes of the rice fields, and water pipes in fluid communication with the irrigation water pipes together with the water level measurement unit; a camera motor configured to adjust the angle of the camera; a valve configured to regulate opening and closing of the irrigation water pipe; a GPS module configured to convert the location of the cultivation observation device into coordinates; a communication unit configured to communicate wired or wirelessly with the user terminal and a water control unit disposed at the water source; and a control unit configured to control the camera, the camera motor, and the valve, and transmit images captured from the camera to the server. Includes.

Also, preferably, the camera motor of the present disclosure includes two servomotors, and the two servomotors are each capable of rotating 180° in mutually orthogonal directions.

Also, preferably, the server of the present disclosure separates the image of the water level measurement unit captured using the camera into RGB channels, obtains a binary image for each channel, and generates R, G, which are binary images for each channel. By calculating the equation below using and B, the brightness of the image converted to a gray image is calculated. At this time, the mathematical formula is as follows.

Also, preferably, the server of the present disclosure calculates the brightness of the gray image as one of 0 to 255, derives a histogram of the number of pixels for each brightness, and uses the peak value extracted from the histogram. Thus, the gray image is corrected using a preset method, and the server uses the corrected gray image to derive the center of gravity of the unlocked portion of the circular marker and uses the center of gravity of the unlocked portion to determine the hidden portion. The y-axis value of is derived, and the water level is calculated using the y-axis value of the hidden part.

In addition, preferably, the cultivation observation device of the present disclosure further includes a water level sensor configured to sense the water level of the irrigation water pipe, and the server records the image of the irrigation water pipe captured by the cultivation observation device. An artificial intelligence model is learned using input data and the water level measured by the water level sensor as output data. When another image of an irrigation water pipe is input to the learned model, the water level of the irrigation water pipe is output.

In addition, preferably, when a control command for opening and closing the valve is input by the user to the user terminal of the present disclosure, the user terminal transmits the control command to the cultivation observation device, and the control unit, The control command is transmitted to the valve, and if it is determined that there is no change in the water level of the irrigation water pipe output by the learning model even though the control command for opening and closing the valve is input by the user to the user terminal, The server determines that the valve is broken and provides a malfunction notification of the valve to the user terminal.

As described above, according to this embodiment, water control is possible automatically or remotely and manually, which has the effect of saving time and labor and lowering costs.

In addition, the cultivation management system according to the present disclosure has the effect of enabling remote water level control by manually or automatically opening and closing the valve using the water level of the rice field observed in real time.

In addition, if there is no change in the water level of the irrigation water pipe even though a command to open and close the valve is generated, it is determined that there is a malfunction of the valve, and the malfunction of the device can be confirmed in real time.

The rice seeding situation during the growing period can be detected through image processing or artificial intelligence, so even workers inexperienced in farming can reduce damage.

By image processing the circular water level measurement unit according to the present disclosure, there is an effect of calculating the accurate water level of the rice field. In particular, the water level measuring unit of the present disclosure has the effect of enabling image reading even in dark situations or when water droplets form.

1 is a configuration diagram of a cultivation management system according to an embodiment of the present disclosure.
Figure 2 is a signal line diagram of a cultivation observation device according to an embodiment of the present disclosure.
Figure 3 is a front view and a side view of a cultivation observation device according to an embodiment of the present disclosure.
Figure 4 is a flowchart of image pre-processing using a cultivation management system according to an embodiment of the present disclosure.
Figure 5 shows the water level measurement unit divided into RGB channels according to an embodiment of the present disclosure.
Figure 6 shows a segmentation process of a binary image according to an embodiment of the present disclosure.
Figures 7 to 9 show pre-processing images of the water level measuring unit according to an embodiment of the present disclosure for each situation.
Figure 10 is a flowchart of a water level calculation method using a preprocessed image according to an embodiment of the present disclosure.
Figures 11 and 12 show examples of the water level calculation method according to Figure 10 of the present disclosure.
Figure 13 is a block diagram showing the configuration of a server according to an embodiment of the present disclosure.

Hereinafter, some embodiments of the present disclosure will be described in detail through illustrative drawings. When adding reference signs to components in each drawing, it should be noted that the same components are given the same reference numerals as much as possible even if they are shown in different drawings. Additionally, in describing the present disclosure, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present disclosure, the detailed description will be omitted.

In describing the components of the embodiment according to the present disclosure, symbols such as first, second, i), ii), a), and b) may be used. These codes are only used to distinguish the component from other components, and the nature, sequence, or order of the component is not limited by the code. In the specification, when a part is said to 'include' or 'have' a certain element, this means that it does not exclude other elements, but may further include other elements, unless explicitly stated to the contrary. .

In the present disclosure, the x-axis refers to a direction horizontal to the ground. Alternatively, when the water level measurement unit 12 is submerged in water, it means a direction parallel to the water surface.

Additionally, in this disclosure, the y-axis refers to a direction perpendicular to the ground. Alternatively, it refers to the direction in which the water level increases when the water level measuring unit 12 is submerged in water.

Additionally, in this disclosure, description is made assuming that the crop is rice. However, note that the present disclosure is not necessarily limited thereto and the types of crops may be different.

Composition of cultivation management system

1 is a configuration diagram of a cultivation management system according to an embodiment of the present disclosure.

Referring to FIG. 1, the cultivation management system 1 according to an embodiment of the present disclosure is composed of a cultivation observation device 10, a water level measurement unit 12, a server 14, and a user terminal 16.

The cultivation observation device 10 is configured to photograph the rice field, the irrigation water pipe 2, and the water pipe (not shown). Here, the irrigation water pipe 2 means a waterway for irrigating the rice field. In addition, a waterway is a narrow passage created to allow water to flow through a rice field, and refers to a waterway configured to communicate fluidly with the irrigation water pipe (2).

Additionally, the cultivation observation device 10 is configured to communicate with the server 14 and the user terminal 16.

The water level measuring unit 12 is placed in the rice field and is arranged so that at least part of the rice field is submerged when the water level of the rice field increases. The water level measurement unit 12 is also photographed by the cultivation observation device 10.

The water level measurement unit 12 includes N circular markers arranged at equal intervals in a direction perpendicular to the surface of the rice field (see Figure 5). At this time, the circular marker may preferably include a luminescent pigment or a luminescent pigment. Because of this, even when the water level measuring unit 12 is photographed in a dark situation, recognition of the circular marker can become easier.

In one example, the water level measuring unit 12 has circular markers arranged at intervals of 4 cm on a 50 cm acrylic plate. At this time, the reason a circular marker is used is because the width or height of the circle is measured consistently no matter what side it is measured from, so water level distortion due to image distortion can be minimized.

Additionally, in the image pre-processing process, the acrylic plate may be painted in a dark color, preferably black, to enable clear image acquisition through contrast.

The server 14 is configured to receive images captured by the cultivation observation device 10, preprocess the received images, and calculate the water level using the preprocessed images. At this time, the water level may be the water level of a rice field or the water level of an irrigation canal. Meanwhile, the image pre-processing method according to the present disclosure will be described in detail with reference to FIGS. 4 to 9. Additionally, the water level calculation method according to the present disclosure will be described in detail with reference to FIGS. 10 to 12.

The user terminal 16 is connected to the cultivation observation device 10 by wire or wirelessly, and may be any electronic device configured to process commands input by the user. In the present disclosure, the user terminal may be, for example, a mobile phone, mobile pad, laptop, PC, etc.

The user terminal 16 is configured to transmit a command input by the user to the cultivation observation device 10.

The user terminal 16 may receive information related to the water level calculated by the server 14 through the cultivation observation device 10 and output it through a display (not shown).

Additionally, the user terminal 16 may receive weather information collected by the weather information collection unit 4 and transmit it to the server 14 through the cultivation observation device 10.

Figure 2 is a signal line diagram of a cultivation observation device according to an embodiment of the present disclosure. Figure 3 is a front view and a side view of a cultivation observation device according to an embodiment of the present disclosure.

2 and 3, the cultivation observation device 10 according to an embodiment of the present disclosure includes a camera 100, a camera motor 101, a valve 102, a GPS module 103, and a communication unit 104. ), including all or part of the control unit 105 and power supply unit 106.

The camera 100 is configured to photograph the water level measuring unit 12, the irrigation water pipe 2, and the water spigot.

The camera motor 101 is configured to adjust the angle of the camera 100. Specifically, the camera motor 101 is preferably comprised of two servo motors. One of the two servomotors may be configured to rotate the camera 100 by more than 180 degrees in the vertical direction, and the other one may be configured to rotate the camera 100 by more than 180 degrees in the horizontal direction.

Meanwhile, the camera 100 may preferably be a pan-tilt-zoom camera (see FIG. 3). That is, by providing the camera motor 101 inside the camera 100, the camera 100 can be configured to capture surrounding images while rotating only the lens.

The valve 102 is configured to control the opening and closing of the irrigation water pipe 2. The cultivation observation device 10 receives control commands related to valve opening or closing from the server 14 or the user terminal 16 and transmits them to the valve 102 in the form of an electrical signal, thereby opening and closing the valve 102. can be controlled.

At this time, the opening and closing of the valve 102 may be performed by an opening and closing motor (not shown) connected to the valve.

Additionally, the valve 102 and one end of the irrigation water pipe 2 may be connected with a circular plastic pipe (not shown). As a result, even if the size and length of the irrigation water pipe 2 are different depending on the region, topography, farmland clearing period, etc., the valve 102 can be opened and closed in various situations because the valve 102 only needs to close the circular plastic pipe. The action can be performed.

The GPS module 103 is configured to determine the location of the cultivation observation device 10 and the location of the rice field. The GPS module 103 includes a real time clock (RTC) function and can simultaneously acquire location information and time information. The acquired location information and time information can be transmitted from the GPS module 103 to the server 14 or the user terminal 16.

Additionally, the position accuracy of the GPS module 103 preferably has a circular error probability (CEP) of less than 3 m.

The communication unit 104 is configured to communicate with the user terminal 16 and the water control unit 3 by wire or wirelessly. The communication unit 104 receives the user's control command from the user terminal 16 and transmits it to the water control unit 3. Because of this, the user can remotely control the opening and closing of the water gate.

At this time, the communication unit 104 can communicate with the user terminal 16 using WiFi or WLAN. Meanwhile, it should be noted that the present disclosure is not necessarily limited to this, and that the communication unit 104 and the user terminal 16 may adopt an appropriate communication method such as Bluetooth or wired communication.

The control unit 105 controls the camera 100, the camera motor 101, and the valve 102, and transmits the image captured by the camera 100 to the server 14, so that the server 14 calculates the water level. It is configured so that it can be done.

The control unit 105 and the camera 100 can communicate using the TCP/IP method. At this time, the control unit 105 can control shooting by transmitting an electrical signal to the camera 100. Here, controlling the shooting means making shooting decisions, adjusting exposure and sensitivity, etc.

Additionally, the control unit 105 may control the camera 100 and the camera motor 101 so that the camera 100 periodically photographs crops planted in a rice field.

When controlling the camera motor 101 and the valve 102, the control unit 105 uses pulse width modulation (PWM) control. Because of this, the control unit 105 can precisely adjust the rotation angles of the camera motor 101 and the valve 102.

The control unit 105 can communicate with the GPS module 103 using a UART (universal asynchronous receiver/transmitter) method. The control unit 105 collects the location information and time information transmitted by the GPS module 103 and transmits it to the communication unit 104, and the communication unit 104 can transmit it to the user terminal 16 or the server 14. .

The power supply unit 106 is configured to provide power to all or part of the camera 100, camera motor 101, valve 102, GPS module 103, communication unit 104, and control unit 105. Meanwhile, in one embodiment of the present disclosure, the power supply unit 106 may include a solar panel (see FIG. 3).

The power supply unit 106 controls the power supplied from the solar panel, stores it in a built-in battery (not shown), and delivers it to an inverter (not shown). When DC power is converted to AC power by the inverter, power can be supplied to each component of the cultivation observation device 10.

Meanwhile, the present disclosure is not necessarily limited to this, and if a constant power source is provided near the cultivation observation device 10, the power supply unit 106 is connected to the constant power source, and the power supplied from the constant power source is used for cultivation observation. It can be delivered to each component of the device 10.

Image preprocessing methods

Figure 4 is a flowchart of image pre-processing using a cultivation management system according to an embodiment of the present disclosure.

Referring to FIG. 4, the image pre-processing method using the cultivation management system 1 according to the present disclosure includes the following steps.

An image of the water level measurement unit is captured using the camera 100 (S400). The captured image is transmitted from the cultivation observation device 10 to the server 14.

The server 14 separates the captured image into RGB channels and obtains a binary image for each channel (S410). In other words, the distribution of each color can be confirmed by dividing the multi-channel image into the single channels R channel, G channel, and B channel. At this time, the process of obtaining a binary image by separating the R channel, G channel, and B channel is applied by a method preset by the user.

Afterwards, a gray image is obtained according to [Equation 1] using the binary image of each separated single channel (S420).

Looking at [Equation 1], the image pre-processing process according to the present disclosure is not aimed at detailed analysis of data, but is aimed at recognizing circular markers through the RGB distribution of the entire image. Therefore, the B channel is excluded to best recognize the circular marker. Here, the reason for excluding the B channel is that the color of the circular marker is preferably yellow, and the lightest color among the colors of the circular marker is blue.

Using the converted image of the water level measuring unit 12, that is, the image preprocessing has been completed, the server 14 calculates the measured water level (S430).

Meanwhile, the process of calculating the water level will be described in detail later with reference to FIGS. 10 to 12.

Figure 5 shows the water level measurement unit divided into RGB channels according to an embodiment of the present disclosure.

Figure 5(a) is an example of an original image taken in a bright situation. In addition, Figures 5 (b), (c), and (d) are examples in which the captured images of the water level measurement unit 12 are divided into R channel, G channel, and B channel and displayed as binary images.

It can be seen that the B channel image in (d) of Figure 5 is darker compared to (b) and (c) of Figure 5. This is because, as explained previously, the lightest color among the colors of the circular marker is blue. For this reason, if the brightness in each channel is expressed as a number from 0 to 255, that is, binary data, the average of the binary data of the B channel will be the lowest.

Figure 6 shows a segmentation process of a binary image according to an embodiment of the present disclosure.

As described above, the image pre-processing method according to the present disclosure is not aimed at detailed analysis of data, but rather aims at recognizing markers with pixel values through RGB color distribution. Accordingly, among image analysis techniques, the image histogram technique is used.

Figure 6 (a) is a binary image obtained by processing the captured image of the water level measuring unit 12 according to [Equation 1]. Additionally, Figure 6(b) shows a histogram for binary data for the binary image according to Figure 6(a). Specifically, the horizontal axis in (b) of FIG. 6 represents color intensity as a number from 0 to 255, and the vertical axis represents the number of pixels corresponding to each color intensity.

Referring to (b) of Figure 6, two clusters can be identified, which are concentrated at the maximum value. At this time, the peak values (max1, max2) of the two clusters can be confirmed. According to the example in (b) of FIG. 6, it can be seen that a large number of pixels are distributed at points where the color intensity is approximately 20 and 90.

Figure 6(c) shows the centroid of the binary image derived by a preset method using the median value of the peak values (max1, max2) of the two clusters according to Figure 6(b). . At this time, the center of gravity was indicated by a red cross.

Figures 7 to 9 show pre-processing images of the water level measuring unit according to an embodiment of the present disclosure for each situation.

Figure 7 shows the results of image shooting and image pre-processing assuming a situation where water is present in the water level measuring unit 12 according to the present disclosure. Specifically, Figure 7(a) shows an image taken when the water level measuring unit 12 was submerged in water, which is separated into R, G, and B channels.

Figure 7(b) is a binary image derived using each channel according to Figure 7(a).

Figure 7(c) shows the pixel value for color intensity of the binary image according to Figure 7(b) as a histogram.

Figure 7(d) is an image corrected using a preset method using the two peak values according to Figure 7(c).

Figure 7(e) shows the center of gravity of the corrected image according to Figure 7(d) obtained by a preset method. Referring to (e) of FIG. 7, the water level measuring unit 12 according to the present disclosure produces an image in which the background and the circular marker are clearly distinguished and the center of gravity of the circular marker is easily identified even when submerged in water. You can check what is being derived.

For this reason, even in a situation where water forms on the water level measuring unit 12 in an actual rice field, the circular marker can be easily identified by using the water level measuring unit 12 according to the present disclosure and the image preprocessing method according to the present disclosure. There is an advantage.

Figure 8 shows the results of image taking and image pre-processing under the assumption that at least a portion of the water level measuring unit 12 according to the present disclosure is submerged in fresh water. Specifically, Figure 8(a) shows an image taken when the water level measuring unit 12 was submerged in water, which is separated into R, G, and B channels. Referring to (a) of FIG. 8, it can be seen that a portion of the water level measurement unit 12 appears reflected in the water at the bottom of each image.

Figure 8(b) is a binary image derived using each channel according to Figure 8(a).

Figure 8(c) shows the pixel value for color intensity of the binary image according to Figure 8(b) as a histogram.

Figure 8(d) is an image corrected using a preset method using the two peak values according to Figure 8(c). Referring to (d) of FIG. 8, it can be seen that the part of the water level measurement unit 12 submerged in water and the part reflected in the water do not appear in the corrected image.

Figure 8(e) shows the center of gravity of the corrected image according to Figure 8(d) obtained by a preset method. Referring to (e) of FIG. 8, when the water level measuring unit 12 according to the present disclosure is submerged in water, the submerged portion is not identified, and the circular marker of the non-submerged portion and the weight of the circular marker are displayed. It can be confirmed that only the center is derived.

For this reason, even in a situation where at least a portion of the water level measurement unit 12 is submerged in fresh water in an actual rice field, the circular marker can be easily identified by using the water level measurement unit 12 according to the present disclosure and the image preprocessing method according to the present disclosure. There is an advantage to being able to do it.

Figure 9 shows the results of image capturing and image pre-processing assuming that the water level measuring unit 12 according to the present disclosure is placed in an environment without a light source. Specifically, Figure 9(a) shows an image taken when the water level measurement unit 12 was submerged in water, which is separated into R, G, and B channels. Referring to (a) of FIG. 9, it can be seen that the circular marker in each image is only vaguely identifiable.

Figure 9(b) is a binary image derived using each channel according to Figure 8(a).

Figure 9(c) shows the pixel value for color intensity of the binary image according to Figure 8(b) as a histogram.

Figure 9(d) is an image corrected using a preset method using the two peak values according to Figure 9(c). Referring to (d) of FIG. 9, it can be seen that the circular marker is clearly identified in the corrected image compared to the image according to (a) of FIG. 9.

Figure 9(e) shows the center of gravity of the corrected image according to Figure 9(d) obtained by a preset method. Referring to (e) of FIG. 9, it can be seen that the water level measuring unit 12 according to the present disclosure clearly derives the circular marker and the center of gravity of the circular marker in a situation where there is no light source.

For this reason, even if the water level measuring unit 12 is photographed at night in an actual rice field, the advantage is that the circular marker can be easily identified by using the water level measuring unit 12 according to the present disclosure and the image preprocessing method according to the present disclosure. there is.

Water level calculation method

Figure 10 is a flowchart of a water level calculation method using a preprocessed image according to an embodiment of the present disclosure. Figures 11 and 12 show examples of the water level calculation method according to Figure 10 of the present disclosure.

With reference to FIGS. 10 to 12 , the water level calculation method according to the present disclosure will be described in detail.

The process of calculating the water level by preprocessing the image of the water level measuring unit 12 received by the server 14 according to the present disclosure includes the following steps.

One circular marker of the water level measuring unit 12 is divided into equal intervals from 0 to 100 based on the y-axis. At this time, if the y-axis value of the circular marker submerged in water is k, the center of gravity of the area where y≥k is derived (S1000). Here, k is an integer between 0 and 100.

Meanwhile, the process of deriving the center of gravity in the S1000 process uses a regression equation based on the correlation between y _c (y-axis value of the center of gravity) and water depth (S1010). At this time, the regression equation can be derived from the relationship between the k value and y _c by calculating the value of y _c according to each k value in advance when k is between 0 and 100 (see FIG. 11).

Specifically, Figure 11(a) shows the case when k=0. Additionally, Figure 11 (b) shows the center of gravity (C) when k is 25. At this time, the y coordinate of C becomes y _c . Additionally, Figure 11(c) shows when k is 50, and Figure 11(d) shows when k is 75.

Meanwhile, preferably, the regression equation can be derived using first-order, second-order, and third-order polynomials.

The results of regression analysis by order according to first-order, second-order, and third-order polynomials are shown in Figure 12. The horizontal axis of FIG. 12 is the cropping ratio due to immersion in water, that is, the k value, and the vertical axis is the y value of the center of gravity (C) accordingly.

Figure 12 (a) shows y _c calculated according to the actual k value. Figures 12 (b) to (d) compare the y value of the center of gravity derived using first-order, second-order, and third-order regression equations, respectively, and the actual y _c . Referring to Figures 12 (b) to (d), it can be seen that the accuracy (R ² ) increases as the degree of the regression equation increases.

Among the converted images, circular markers that are submerged and circular markers that are not submerged and at least partially submerged in water are distinguished (S1020). For example, if three of the circular markers arranged in a row of seven are submerged, one is partially submerged, and the remaining three are not submerged, then four are submerged and three are not submerged. Classify as not.

Afterwards, when it is confirmed that the number of circular markers submerged in water is three and that the circular marker partially submerged is the fourth circular marker, the server 14 uses this to calculate the water level. Specifically, the y value of the center of gravity of the remaining part of the locked part of the fourth circular marker is derived. Afterwards, the height of the submerged portion is determined using the regression equation derived in step S1010. The final water level is calculated by adding the submerged height of the fourth circular marker to the height of the three circular markers fully submerged in water (S1030).

According to the present disclosure, there is an advantage that the water level of the rice field can be accurately calculated through a simple calculation process in steps S1000 to S1030.

Configuration of learning model and prediction method using it

Figure 13 is a block diagram showing the configuration of a server according to an embodiment of the present disclosure.

Referring to FIG. 13, the server 14 according to the present disclosure includes an AI model construction unit 140, an irrigation water pipe image input unit 141, an irrigation water pipe water level input unit 142, a crop image input unit 143, It includes all or part of the crop tillering status input unit 144, the crop variety and transplanting date input unit 145, and the crop heading status input unit 146.

The AI model construction unit 140 uses the image of the irrigation water pipe 2 captured by the camera 100 of the cultivation observation device 10 as input data, and uses the water level sensor (not shown) provided in the irrigation water pipe 2 as input data. An artificial intelligence model is learned using the water level information measured by the city as output data. At this time, the image of the irrigation water pipe 2 is input to the irrigation water pipe image input unit 141. Additionally, the water level of the irrigation water pipe 2 measured by the water level sensor is input to the irrigation water pipe water level input unit 142.

When an image of an irrigation water pipe (2) different from the learned image is input to the learned artificial intelligence model, the water level corresponding to the image can be derived.

The AI model construction unit 140 according to the present disclosure has the advantage of automatically measuring the water level when only the cultivation observation device 10 is installed in the irrigation water pipe 2 that is not equipped with a water level sensor.

In addition, the AI model construction unit 140 learns another artificial intelligence model using the image of the crop captured by the cultivation observation device 10 as input data and the tillering status of the crop as output data. At this time, the image of the crop is input into the crop image input unit 143. Additionally, the tillering status of the crop is input into the tillering status input unit 144 of the crop.

At this time, if an image of a crop different from the learned image is input to another learned artificial intelligence model, a tillering situation corresponding to the image may be derived.

The AI model construction unit 140 according to the present disclosure has the advantage of being able to determine the harvest time through the automatically output tillering status, even if an unskilled worker has no knowledge of the tillering status and tillering stage of the crop.

In addition, the AI model construction unit 140 learns another artificial intelligence model using the image of the crop and the crop variety captured by the cultivation observation device 10 as input data and the transplanting date of the crop as output data. . At this time, the image of the crop is input into the crop image input unit 143. In addition, the crop variety and crop transplant date are input into the crop variety and transplant date input unit 145.

At this time, when an image of a crop and a crop variety different from the learned image are input to another learned artificial intelligence model, the transplanting date corresponding to the image is output.

As the preferred transplanting date of crops is output by the AI model construction unit 140 according to the present disclosure, even an unskilled worker without knowledge of the transplanting time can determine the transplanting time through the automatically output transplanting date. There is.

In addition, the AI model construction unit 140 learns another artificial intelligence model using the image of the crop captured by the cultivation observation device 10 as input data and the heading situation of the crop as output data. At this time, the image of the crop is input into the crop image input unit 143. Additionally, the heading status of crops is input into the crop heading status input unit 146. Here, the heading situation of the crop includes the heading stage in the paddy field.

At this time, when an image of a crop different from the learned image is input to another learned artificial intelligence model, a water harvester corresponding to the image is output.

As the AI model construction unit 140 according to the present disclosure outputs the desired planting time of the crop, even an unskilled worker without knowledge of the planting machine can know the name and lateness of the planting time through the automatically output heading date. There is an advantage in being able to respond appropriately.

Meanwhile, when an image of a crop according to the present disclosure is captured and the image is stored, it is preferable that the location information and time information collected by the GPS module 103 are stored together.

Additionally, the artificial intelligence model may be MLP, CNN, RNN, ANN, etc.

Meanwhile, in addition to the artificial intelligence learning model of the server 14, the growth degree of crops can be analyzed using a rule-base algorithm. For this purpose, the cultivation observation device 10 periodically photographs crops planted in the rice field.

The captured image of the crop is transmitted to the server 14 along with the location information and time information collected by the GPS module 103.

The server 14 calculates the leaf area using a preset method using the photographed image of the crop. Additionally, the server 14 can use the calculated leaf area to analyze the degree of growth according to preset standards.

At this time, as the variety of the crop planted in the rice field of the server 14 is input, the shooting time, growth degree, and variety are stored in the server 14 as growth data.

The server 14 can organize growth data by year and output it through the user terminal 16. At this time, it is desirable that annual growth data be output in the form of a table or graph.

Additionally, by using the cultivation management system 1 according to the present disclosure, the user can determine whether the valve 102 is malfunctioning.

Specifically, when the user inputs a control command to open and close the valve 102 in the user terminal 16, the input control command is transmitted to the valve 102 through the communication unit 104 of the cultivation observation device 10. do.

At this time, the server 14 can calculate the water level in real time using the image of the irrigation water pipe 2 captured by the cultivation observation device 10. On the other hand, if it is determined that the water level of the irrigation water pipe (2) does not change even after a predetermined time has elapsed after the valve 102 control command is input by the user, the server 14 determines that the valve 102 is broken. judge. If it is determined that the valve 102 is malfunctioning, the server 14 may provide a valve malfunction notification to the user through the user terminal 16.

Because of this, there is an advantage that the user can determine whether the valve 102 is broken in real time.

The above description is merely an illustrative explanation of the technical idea of the present embodiment, and those skilled in the art will be able to make various modifications and variations without departing from the essential characteristics of the present embodiment. Accordingly, the present embodiments are not intended to limit the technical idea of the present embodiment, but rather to explain it, and the scope of the technical idea of the present embodiment is not limited by these examples. The scope of protection of this embodiment should be interpreted in accordance with the claims below, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of rights of this embodiment.

1: Cultivation management system
10: Cultivation observation device
100: Camera
101: Camera motor
102: valve
103: GPS module
104: Department of Communications
105: control unit
106: Power supply unit
12: Water level measurement unit
120: Circular marker
14: Server
16: User terminal
2: Irrigation water pipe
3: Hydrological Control Department
4: Weather information collection department

Claims (7)

A water level measuring unit including N circular markers arranged at equal intervals in a direction perpendicular to the surface of the paddy field;
A cultivation observation device configured to photograph the water level measurement unit;
A server configured to receive the image captured by the cultivation observation device and calculate the water level of the rice field; and
It includes a user terminal configured to receive the water level calculated by the server through the cultivation observation device and transmit a control command to the cultivation observation device,
When the water level measuring unit is submerged in water, M circular markers among the N circular markers are submerged in water, and a portion of the M+1 circular marker is submerged,
The server derives the center of gravity of the unlocked part of the M+1th circular marker, derives the y-axis value of the hidden part using the center of gravity of the unlocked part, and calculates the y-axis value of the hidden part. Calculate the water level using
The cultivation observation device is,
A camera configured to photograph crops grown in the rice field, irrigation water pipes of the rice fields, and water pipes in fluid communication with the irrigation water pipes together with the water level measuring unit;
It includes a water level sensor configured to sense the water level of the irrigation water pipe,
The server learns an artificial intelligence model using the image of the irrigation water pipe captured by the camera as input data and the water level of the irrigation water pipe measured by the water level sensor as output data,
When another image of an irrigation water pipe is input to the learned model, the water level of the irrigation water pipe is output.
Cultivation management system.
According to paragraph 1,
a camera motor configured to adjust the angle of the camera;
a valve configured to regulate opening and closing of the irrigation water pipe;
a GPS module configured to convert the location of the cultivation observation device into coordinates;
a communication unit configured to communicate wired or wirelessly with the user terminal and a water control unit disposed at the water source; and
a control unit configured to control the camera, the camera motor, and the valve, and transmit images captured from the camera to the server; Including,
Cultivation management system.
According to paragraph 2,
The camera motor includes two servomotors,
The two servomotors are each capable of rotating 180° in mutually orthogonal directions,
Cultivation management system.
According to paragraph 2,
The server is,
Separate the image of the water level measurement unit taken using the camera into RGB channels, and obtain a binary image for each channel,
By calculating the equation below using the binary images R, G and B for each channel, the brightness of the image converted to a gray image is calculated,

Cultivation management system.
According to paragraph 4,
The server is,
The brightness of the gray image is calculated as one of 0 to 255.
Deriving a histogram of the number of pixels for each of the brightnesses,
Correcting the gray image using a preset method using the peak value extracted from the histogram,
The server is,
Using the corrected gray image, the center of gravity of the unlocked part of the circular marker is derived, the y-axis value of the hidden part is derived using the center of gravity of the unlocked part, and the y-axis value of the hidden part is used. to calculate the water level,
Cultivation management system.
delete According to paragraph 2,
When a control command for opening and closing the valve is input to the user terminal by the user,
The user terminal transmits the control command to the cultivation observation device,
The control unit transmits the control command to the valve,
If it is determined that there is no change in the water level of the irrigation water pipe output by the learned model even though a control command for opening and closing the valve is input to the user terminal by the user,
The server determines that the valve is broken and provides a malfunction notification of the valve to the user terminal.
Cultivation management system.

Images