TWI659192B - Intelligent cultivation apparatus - Google Patents

Abstract

The invention discloses an intelligent farming device for solving problems such as poor farming efficiency, excessive time and labor costs, and environmental pollution. The system includes: a vehicle body, which is pivotally connected to a plurality of wheels by a frame body; a driving device combined with the frame body; an agricultural component provided on the driving device; an image capturing unit provided on the frame A body; and a processing unit, which is disposed on the frame body and is electrically connected to the driving device and the image capturing unit, the processing unit controls the image capturing unit to shoot toward a soil to generate a soil image, the processing unit Based on the results obtained from analyzing the soil image, the driving device is driven to control the farming component to operate.

Description

Intelligent farming device and control method thereof

The invention relates to an intelligent farming device and a control method thereof, in particular to an intelligent farming device and a control method thereof for performing farming operations based on image recognition.

In the implementation of organic cultivation, the control of weeds plays a very important role, because weeds will compete with crops for nutrients and moisture in the soil, and further occupy the living space of crops, which will lead to reduced crop yield and quality. Therefore, in order to prevent weeds from growing, the conventional weeding method uses a human-powered lawnmower to carry out weeding operations, but it has problems such as poor weeding efficiency and high time and labor costs, which are no longer suitable for the needs of modern agricultural development. For this reason, another conventional weeding method is to spray chemicals to carry out weeding operations, which can improve the weeding efficiency and reduce the time and labor costs. However, due to factors such as excessive weed species and resistance, it cannot be permanent and Comprehensive solution to the problem of weeds. In addition, when the chemical agent is improperly used, it may easily cause environmental pollution such as soil and water, and even cause poisoning of the applicator.

In view of this, it is necessary to provide an intelligent farming device to solve the aforementioned problem of weeding methods.

In order to solve the above problems, an object of the present invention is to provide an intelligent farming device that can analyze and execute corresponding farming operations with image recognition technology to improve farming efficiency.

A secondary object of the present invention is to provide an intelligent farming control method using the above-mentioned device.

The following directional or similar terms of the present invention, such as "front", "rear", "up (top)", "down (bottom)", "inside", "outside", "side", etc. With reference to the directions of the accompanying drawings, each directivity or its approximate terms are only used to assist in explaining and understanding the embodiments of the present invention, and are not intended to limit the present invention.

The intelligent farming device of the present invention comprises: a vehicle body, a plurality of wheels being pivotally connected by a frame body; a driving device combined with the frame body; an agricultural component provided in the driving device; an image capturing unit Is provided on the frame body; a sprinkler member is combined with the frame body; and a processing unit is provided on the frame body and is electrically connected with the driving device, the image capturing unit and the sprinkler member The processing unit controls the image capturing unit to shoot toward a soil to generate a soil image. The processing unit analyzes the soil image through a fuzzy controller to determine the surface humidity of the soil and generate a watering required for crops. A pulse modulation signal is obtained according to the watering amount conversion, and the pulse modulation signal is used to control the sprinkler member to water the crops.

The intelligent farming control method of the present invention comprises: shooting an image capturing unit toward a soil to generate a soil image; converting the soil image from an RGB color space to an HSV color space by a processing unit to generate an HSV image; The processing unit inputs a threshold value to the HSV image to perform a binarization process so as to generate a grayscale image; the processing unit sets a plurality of regions of interest in the grayscale image, and the plurality of regions of interest include a first A region of interest, a second region of interest, and a third region of interest. The second region of interest includes crop images. The first region of interest and the third region of interest may be located in the second sense region. Both sides of the region of interest; the processing unit masks the second region of interest with a mask, and performs a connected component labeling on the first region of interest and the third region of interest to obtain two non-crop region images ; And the processing unit discriminates whether weeds exist in the image of the non-crop area, and if the discrimination result is yes, then controls a corresponding driving device to drive a weeder to rake weeds

According to this, the intelligent farming device of the present invention can perform corresponding farming operations through image recognition, has the effect of improving farming efficiency, and can avoid excessive irrigation or irrigation Insufficient, it has the effect of improving the survival rate of crops.

Wherein, the agricultural component is a weeder, and the processing unit analyzes whether there is weed in the soil image. If the analysis result is yes, the driving device is driven to drive the weeder to rake the weed in the soil. . In this way, it has the effect of reducing labor and time costs.

Wherein, the agricultural component is a sprayer, and the processing unit confirms whether there is a crop in the soil image. If the result of the confirmation is no, the driving device is driven to drive the sprayer to spray or fertilize the crop. In this way, it has the effect of improving the range of use.

The sprinkler member has a water tank, a water pump and a water spray member. The water tank is provided in the frame body. The water pump is disposed in the water tank. One end of the water spray member is connected to the water pump and the other end It is arranged adjacent to the driving device, and the processing unit controls the voltage value of the pumping pump according to the pulse modulation signal to extract a water body equivalent to the watering amount from the water tank to the water spraying member. In this way, the structure of the sprinkler member is simple, and has the functions of convenient assembly and convenient maintenance.

The intelligent farming device of the present invention may further include a sensing module, which is coupled to the processing unit and senses the internal humidity of the soil with a humidity sensing unit to generate a humidity value. The processing The unit analyzes the humidity value and the soil image through the fuzzy controller to determine the watering amount required by the crop. In this way, it has the effect of precisely controlling the amount of water.

Wherein, the driving device drives a first gear of a transmission member with a driving member to rotate, and causes the first gear to synchronously rotate a second gear with a transmission belt, and a plate is fixedly connected to one side of the second gear. One end of the plate is pivotally connected to one end of a connecting member. The connecting member is slidably disposed on a limit rod of the frame body through a long groove. The agricultural component is detachably coupled to the other end of the connecting member. The driving device is electrically connected to a power supply by the driving member. In this way, the manufacturing process of the driving device is simple and has the effect of improving production efficiency.

The processing unit masks the first region of interest and the third region of interest with the mask, and then performs an area scan and the connected component marking on the second region of interest to obtain a crop region Image, the processing unit confirms whether there is a crop in the image of the crop area Here, if the result of the confirmation is yes, the corresponding sprinkler member is controlled to water the crops. In this way, excessive irrigation or insufficient irrigation can be avoided, and the system has the effect of improving the survival rate of crops.

The processing unit uses a damp state of the soil surface to which the soil image belongs as a first input variable of a fuzzy algorithm to generate a pulse modulation signal, and the processing unit controls the sprinkler with the pulse modulation signal. The voltage value of a water pump of a water component. In this way, it has the effect of precisely controlling the amount of water.

The processing unit uses a humidity sensing unit to obtain the internal humidity of the soil to generate a humidity value, and uses the humidity value as a second input variable of the fuzzy algorithm. In this way, the system has the effect of more precisely controlling the amount of water.

The processing unit sets the RGB value of the part of the soil image relative to the crop area image to 0, and calculates an average of an R-channel image, a G-channel image, and a B-channel image in the soil image. And obtain an average value of the R-channel image, the G-channel image, and the B-channel image, and the processing unit judges and obtains the wet state of the soil surface to which the soil image belongs according to each of the average values and the average value. In this way, the system has the effect of improving the accuracy of determining the wet state.

The threshold is a fixed threshold. The processing unit obtains an H-channel image from the HSV image and enters the threshold to perform a binarization process on the H-channel image. In this way, weeds can be effectively judged in various different periods, which has the effect of improving the recognition rate of weeds.

The threshold is an adaptive threshold. The processing unit obtains an H-channel image, an S-channel image, and a V-channel image from the HSV image, and calculates the H-channel image, the S-channel image, and An average value of the V-channel image is used to calculate whether the average value of the H-channel image is greater than the average value of the S-channel image. If the calculation result is yes, the processing unit enters the threshold to perform binarization processing on the S-channel image. . In this way, the crops can be judged according to different time periods, which has the effect of improving the identification rate of crops.

Wherein, if the calculation result is no, the processing unit inputs the threshold value for the H channel. The image is binarized. In this way, the system has the effect of improving the recognition rate of crops.

The processing unit performs a morphological process on the grayscale image. In this way, in order to eliminate noise, it has the effect of improving the accuracy of subsequent image processing.

〔this invention〕

1‧‧‧ body

11‧‧‧frame body

12‧‧‧ Wheel

13‧‧‧ limit lever

2‧‧‧Drive

21‧‧‧Driver

22‧‧‧ Transmission Parts

221‧‧‧first gear

222‧‧‧Drive Belt

223‧‧‧Second Gear

23‧‧‧ Plate

24‧‧‧Connector

241‧‧‧long slot

25‧‧‧ Power Supply

3‧‧‧ farming components

3a‧‧‧ Weeder

3b‧‧‧ sprayer

4‧‧‧Image capture unit

5‧‧‧ processing unit

6‧‧‧ sprinkler

61‧‧‧Water tank

62‧‧‧Pump pump

63‧‧‧Water spray

7‧‧‧ Sensor Module

S1‧‧‧Image shooting steps

S2‧‧‧Image processing steps

S21‧‧‧Image Compression Procedure

S22‧‧‧color code conversion steps

S23‧‧‧ Binarization step

S24‧‧‧Feature Enhancement Steps

S25‧‧‧Image analysis steps

Figure 1: A perspective view of the intelligent farming device of the present invention.

Fig. 2: Rear view taken along line A-A of Fig. 1.

Figure 3: A side view of the intelligent farming device of the present invention.

Figures 4 to 7: Operation situation diagrams of the driving device of the intelligent farming device of the present invention.

FIG. 8 is a method flowchart of the intelligent farming control method of the present invention.

In order to make the above and other objects, features, and advantages of the present invention more comprehensible, the following describes the preferred embodiments of the present invention in detail with the accompanying drawings as follows: Please refer to FIG. 1, It is a preferred embodiment of the intelligent farming device of the present invention, and includes a vehicle body 1, at least one driving device 2, an agricultural component 3, an image capturing unit 4, and a processing unit 5. The at least one driving device 2 The image capturing unit 4 and the processing unit 5 are respectively installed on the vehicle body 1, the farming component 3 is combined with the at least one driving device 2, and the at least one driving device 2 and the image capturing unit 4 are electrically This processing unit 5 is connected.

The size and style of the vehicle body 1 are not limited in the present invention. The vehicle body 1 may include a frame body 11. The frame body 11 is pivotally connected to a plurality of wheels 12 and has a limit rod 13.

Please refer to FIG. 2, the at least one driving device 2 is coupled to the frame body 11. In this embodiment, the number of the at least one driving device 2 may be three. Each of the driving devices 2 drives a first gear 221 of a transmission member 22 with a driving member 21 to rotate, so that the first gear 221 drives a second gear 223 with a transmission belt 222 to rotate synchronously. The second gear 223 A plate 23 is fixed on one side, and one end of the plate 23 is pivotally connected to one end of a connecting member 24. The connecting member 24 is slidably disposed on the limiting rod 13 with a long groove 241. Wherein, the at least one driving device 2 is electrically connected to a power supply 25 by the driving member 21, the driving member 21 may be a servo motor or a stepping motor, and the transmission belt 222 may be a chain or a gear belt, but not This is the limit.

The farming component 3 can be separately coupled to the other end of the connecting member 24. In this embodiment, the agricultural component 3 may include one or more of a weeder 3a (such as a grazing rake), a sprayer 3b, or a seeder. The sprayer 3b is used for spraying or fertilizing crops, and the seeder is used for seeding. It can be understood by those with ordinary knowledge in the technical field to which the present invention belongs, and it will not be repeated here.

The image capturing unit 4 is disposed on the frame body 11. Preferably, the image capturing unit 4 can be disposed below the frame body 11. The image capturing unit 4 is used for shooting towards a soil to generate a soil image. For example, the image capturing unit 4 may be a CCD camera lens, but is not limited thereto.

The processing unit 5 is disposed on the frame body 11 and is electrically connected to the at least one driving device 2 and the image capturing unit 4. The processing unit 5 may be any processor having functions such as data processing, image recognition processing, signal generation, and control, for example, a microprocessor (MCU). The processing unit 5 drives the at least one driving device 2 to control the agricultural component 3 to act according to a result obtained by analyzing the soil image. Specifically, when the agricultural component 3 is a weeder 3a, the processing unit 5 analyzes whether there is weed in the soil image, and if the analysis result is yes, the at least one driving device 2 is driven to drive the weeder 3a rakes the weeds in the soil; if the analysis result is no, it is not necessary to drive the at least one driving device 2 to operate. When the farming component 3 is a sprayer 3b, the processing unit 5 confirms whether there is a crop in the soil image, and if the confirmation result is yes, it is not necessary to drive the at least one driving device 2 to act; if the confirmation result is If not, the at least one driving device 2 is driven to drive the sprayer 3b to spray or fertilize the crop. When the farming component 3 is a seeder, the processing unit 5 drives the at least one drive The moving device 2 drives the planter to sow the soil.

Please refer to FIG. 3 together. The intelligent farming device of the present invention may further include a sprinkler member 6 which is coupled to the frame body 11 and is electrically connected to the processing unit 5. The sprinkler The component 6 may include a water tank 61, a pumping pump 62, and a water spraying member 63. The water tank 61 is disposed in the frame body 11. The water pumping pump 62 is disposed in the water tank 61. One end of the water spraying member 63 is connected to the The other end of the suction pump 62 may be disposed adjacent to the at least one driving device 2. The processing unit 5 can analyze the soil image through a fuzzy controller to determine the surface humidity of the soil, so as to generate a watering amount required for the crop. The processing unit 5 then obtains a pulse modulation signal (PWM) according to the watering amount conversion, and controls the voltage value of the pumping pump 62 with the pulse modulation signal, and extracts from the water tank 61 equal to the watering amount The water body to the water spraying member 63 to water the crops in the soil.

Please refer to FIG. 2 again. The intelligent farming device of the present invention may further include a sensing module 7 which is coupled to the processing unit 5 and senses the humidity with a humidity sensing unit. The internal humidity of the soil to produce a humidity value. The processing unit 5 can further analyze the humidity value and the soil image with the fuzzy controller to determine the watering amount required for the crops in the soil. The sensing module 7 can transmit the humidity value to the processing unit 5 by a transmission unit. For example, but not limited to, the transmission unit can be a Zigbee wireless transmission module. In addition, the sensing module 7 may be set corresponding to each crop, or may be set corresponding to the growth range of several crops, which is not limited in the present invention.

Please refer to FIGS. 4 to 7. According to the foregoing structure, when the processing unit 5 analyzes that weeds are present in the soil image, the processing unit 5 controls the driving member 21 to drive the transmission member 22 to make the transmission The member 22 drives the plate member 23 to rotate. In this embodiment, the plate member 23 is rotated 90 degrees clockwise. When the plate member 23 rotates, the connecting member 24 is moved along the limit rod 13. Among them, FIG. 4 is a state when the plate 23 is not rotated, and FIGS. 5 to 7 are states when the plate 25 is rotated 90 degrees, 180 degrees, and 270 degrees, respectively. When the plate 23 is rotated 90 degrees, This will drive the connecting member 24 to move backward. When the plate member 23 is rotated 180 degrees, the connecting member 24 is driven to move downward and forward, so that the agricultural component 3 is inserted into the soil. When the plate member 23 is rotated 270 degrees, the connecting member 24 is further moved upward and forward, so that the agricultural component 3 is protruded from the soil to complete the weeding and weeding action.

Please refer to FIG. 8, which is a preferred embodiment of the control method of the intelligent farming device of the present invention, and can be applied to the above-mentioned intelligent farming device. The control method includes: an image capturing step S1 and an image processing step S2.

The image capturing step S1 can be performed by using an image capturing unit 4 toward a soil to generate a soil image. In this embodiment, the soil image is a color image.

The image processing step S2 may include an image compression step S21, a color code conversion step S22, a binarization step S23, a feature enhancement step S24, and an image analysis step S25. In detail, the image compression step S21 is to compress the soil image to improve subsequent image processing efficiency. In this embodiment, the compression ratio of the soil image is 75%. The color code conversion step S22 can convert the soil image from the RGB color space to the HSV color space by a processing unit 5 to generate an HSV image. The conversion from RGB color space to HSV color space can be understood by those having ordinary knowledge in the technical field to which the present invention pertains, and details are not described herein.

In the binarization step S23, the processing unit 5 inputs a threshold value to the HSV image to perform binarization processing, so as to generate a grayscale image. In this embodiment, when the threshold is a fixed threshold, the processing unit 5 obtains an H-channel image from the HSV image, and inputs the threshold to perform a binarization process on the H-channel image. In addition, when the threshold is an adaptive threshold, the processing unit 5 obtains an H-channel image, an S-channel image, and a V-channel image from the HSV image, and calculates the H-channel image and the S, respectively. An average of the channel image and the V-channel image. The processing unit 5 calculates whether the average value of the H-channel image is greater than the average value of the S-channel image. If the calculation result is yes, the processing unit 5 uses the threshold value for the S-channel. The image is binarized. The processing unit 5 recalculates the threshold to generate a new threshold, and iteratively calculates the new threshold with the old threshold until the new threshold converges; if the calculation result is no, the processing unit 5 Enter this threshold to perform binarization on the H-channel image. In this embodiment, the initial value of the threshold is set to 0, and the new threshold is the average value of the V-channel image, multiplied by the sum of the image area of the weeds and the crops in the grayscale image. If the new threshold is subtracted from the old threshold by not less than 0.0001, the processing unit 5 re-binarizes the selected channel image with the new threshold.

In the feature enhancement step S24, a morphological process (Morphology) can be performed on the grayscale image by the processing unit 5. In this embodiment, the morphological process can be Erosion or Dilation.

In the image analysis step S25, a plurality of Regions of Interest (ROI) can be set in the grayscale image by the processing unit 5. The number of the region of interest is set corresponding to the number of the at least one driving device 2. In this embodiment, the number of the region of interest may be three, that is, a first region of interest, a second region of interest Area and a third area of interest. The second region of interest includes crop images, and the first region of interest and the third region of interest may be located on two sides of the second region of interest, respectively. The processing unit 5 masks the second region of interest with a mask, and performs a connected component labeling on the first region of interest and the third region of interest to obtain two non-crop region images, In this embodiment, the connected component label may be an 8-connected component label. The processing unit 5 discriminates whether weeds exist in the two non-crop area images by a maximum area method. If the discrimination result is yes, the corresponding driving device 2 is controlled to drive a weeder 3a to rake weeds.

The processing unit 5 further masks the first region of interest and the third region of interest with the mask, and then performs an area scan and the connected component marking on the second region of interest to obtain a crop region. image. The processing unit 5 confirms whether there is a crop in the crop area image by using the maximum area method, and if the confirmation result is yes, the corresponding driving device 2 can be controlled Drive a sprayer 3b to spray or fertilize the crop, or control a sprinkler member 6 to water the crop; if the result of the confirmation is no, it means that the image of the crop area does not include crops, and the process The unit 5 reconfirms whether weeds exist in the crop area. If the result of the confirmation is yes, the corresponding driving device 2 can be controlled to drive the weeder 3a to rake weeds in the second region of interest.

Wherein, the processing unit 5 uses a damp state of the soil surface to which the soil image belongs as a first input variable of a fuzzy algorithm to generate a pulse modulation signal (PWM), and the processing unit 5 adjusts the pulse with the pulse. The variable signal controls the voltage value of a pumping pump 62 of the sprinkler member 6. Specifically, the processing unit 5 sets the RGB value of the portion of the soil image relative to the crop area image to 0, and the processing unit 5 calculates an R-channel image, a G-channel image, and a An average value of the B-channel image and an average value of the R-channel image, the G-channel image, and the B-channel image. The processing unit 5 determines the wet state of the soil surface to which the soil image belongs according to each of the average values and the average values.

Preferably, the processing unit 5 obtains the internal humidity of the soil by a humidity sensing unit to generate a humidity value, and uses the humidity value as a second input variable of the fuzzy algorithm. The fuzzy algorithm sets the fuzzy statement of the wet state (image) of the soil surface as wet, mid, and dry, and sets the humidity value (sensor) obtained by the humidity sensing unit respectively. The fuzzy sentences of the output pulse modulation signal are small, mid, and large. The fuzzy rule base of the fuzzy algorithm can be shown in Table 1 below:

In summary, the intelligent farming device and its control method of the present invention can perform corresponding farming operations through image recognition, and has the effect of improving farming efficiency. In addition, when it is determined that it is a crop, relative spraying, fertilizing, or watering has the effect of further improving farming efficiency. Furthermore, the fuzzy controller analyzes and obtains the watering amount required for the crop, and converts to obtain the pulse modulation signal for controlling the voltage value of the pumping pump according to the watering amount conversion, which has the purpose of avoiding excessive irrigation or insufficient irrigation, etc. efficacy.

Although the present invention has been disclosed using the above-mentioned preferred embodiments, it is not intended to limit the present invention. Anyone skilled in the art can make various changes and modifications to the above embodiments without departing from the spirit and scope of the present invention. The technical scope protected by the invention, so the scope of protection of the present invention shall be determined by the scope of the appended patent application.

Claims (15)

An intelligent farming device includes: a vehicle body pivotally connected to a plurality of wheels by a frame body; a driving device coupled to the frame body; an agricultural component disposed on the driving device; an image capturing unit, To the frame body; a sprinkler member coupled to the frame body; and a processing unit disposed on the frame body and electrically connected to the driving device, the image capturing unit and the sprinkler member, the The processing unit controls the image capturing unit to shoot toward a soil to generate a soil image. The processing unit analyzes the soil image through a fuzzy controller to determine the surface humidity of the soil and generate a watering amount required for the crop. A pulse modulation signal is obtained according to the watering amount conversion, and the pulse modulation signal is used to control the sprinkler component to water the crops. The intelligent farming device according to item 1 of the scope of patent application, wherein the farming component is a weeder, the processing unit analyzes whether there is weed in the soil image, and if the analysis result is yes, the driving device is driven The weeder is driven to weed the soil. According to the intelligent farming device described in the first item of the patent application scope, wherein the farming component is a sprayer, the processing unit confirms whether there is a crop in the soil image, and if the confirmation result is no, the driving device is driven Drive the sprayer to spray or fertilize crops. The intelligent farming device according to item 1 of the scope of patent application, wherein the sprinkler member has a water tank, a water pump and a water spraying member, the water tank is provided on the frame body, and the water pump is provided on the water tank Inside, one end of the water spray member is connected to the pump, and the other end is disposed adjacent to the driving device. The processing unit controls the voltage value of the pump according to the pulse modulation signal, so as to be extracted from the water tank equivalent to the watering. Amount of water body to the water spray member. The intelligent farming device described in item 1 of the patent application scope further includes a sensing module, which is coupled to the processing unit and senses the internal humidity of the soil with a humidity sensing unit to generate a Humidity value, the processing unit analyzes the humidity value and the soil image through the fuzzy controller to determine the amount of watering required by the crop. The intelligent farming device according to item 1 of the scope of patent application, wherein the driving device drives a first gear of a transmission member to rotate with a driving member, and causes the first gear to synchronously rotate a second gear with a driving belt, A plate is fixedly connected to one side of the second gear, one end of the plate is pivotally connected to one end of a connecting member, and the connecting member is slidably disposed on a limit rod of the frame body through a long groove. The component is detachably coupled to the other end of the connecting member, wherein the driving device is electrically connected to a power supply by the driving member. An intelligent farming control method uses the intelligent farming device described in any one of items 1 to 6. The control method includes: shooting an image capturing unit toward a soil to generate a soil image; and using a processing unit to The soil image is converted from the RGB color space to the HSV color space to generate an HSV image; the processing unit inputs a threshold value to the HSV image to perform a binarization process so as to generate a grayscale image; the processing unit generates a grayscale image; A plurality of regions of interest are set in the first-order image. The plurality of regions of interest include a first region of interest, a second region of interest, and a third region of interest. The second region of interest includes a crop image. A region of interest and the third region of interest may be located on both sides of the second region of interest, respectively; the processing unit masks the second region of interest with a mask, and The third region of interest performs a connected component labeling to obtain an image of the second non-crop area; and the processing unit discriminates whether there is weed in the non-crop area image, and if the discrimination result is , A drive control means to drive a corresponding rake herbicidal weed control. The intelligent farming control method according to item 7 of the scope of patent application, wherein the processing unit masks the first region of interest and the third region of interest with the mask, and then executes a second region of interest The area scanner and the connected component mark to obtain a crop area image, the processing unit confirms whether there is a crop in the crop area image, and if the result of the confirmation is yes, controls a corresponding sprinkler member to water the crop . The intelligent farming control method according to item 8 of the scope of patent application, wherein the processing unit uses a wet state of the soil surface to which the soil image belongs as a first input variable of a fuzzy algorithm to generate a pulse modulation Signal, the processing unit then uses the pulse modulation signal to control a voltage value of a pumping pump of the sprinkler member. The intelligent farming control method according to item 9 of the scope of patent application, wherein the processing unit obtains the internal humidity of the soil with a humidity sensing unit to generate a humidity value, and the humidity value is used as a first part of the fuzzy algorithm. Two input variables. The intelligent farming control method according to item 9 of the scope of the patent application, wherein the processing unit sets the RGB value of the portion of the soil image relative to the image of the crop area to 0, and calculates an R in the soil image An average value of the channel image, a G-channel image, and a B-channel image, and obtaining an average value of the R-channel image, the G-channel image, and the B-channel image, and the processing unit judges and obtains the average value and the average value The wet state of the soil surface to which this soil image belongs. The intelligent farming control method as described in item 7 of the scope of patent application, wherein the threshold value is a fixed threshold value, the processing unit obtains an H-channel image from the HSV image, and inputs the threshold value to the H-channel image Binarization is performed. The intelligent farming control method according to item 7 of the scope of patent application, wherein the threshold is an adaptive threshold, and the processing unit obtains an H-channel image, an S-channel image, and a V-channel from the HSV image, respectively. Image, and calculate an average value of the H-channel image, the S-channel image, and the V-channel image respectively, and calculate whether the average value of the H-channel image is greater than the average value of the S-channel image. If the calculation result is yes, the process The unit inputs the threshold to perform binarization processing on the S-channel image. The intelligent farming control method according to item 13 of the scope of patent application, wherein if the calculation result is no, the processing unit inputs the threshold value to perform binarization processing on the H-channel image. The intelligent farming control method according to item 7 of the scope of patent application, wherein the processing unit performs a morphological process on the grayscale image.