Leaf surface frosting amount detection method, crop defrosting system and crop defrosting method

Abstract

The invention discloses a method for detecting the frosting amount of a leaf surface. The method comprises the first step of preparation and the second step of continuous detection. Controlling the gray-scale camera to shoot a gray-scale image of the selected typical blade once every X minutes, and counting SW and SS by the electric control device after shooting every time; when SW / S1 is greater than 0.95, the leaf surface is basically frostless, and at the moment, the electric control device does not calculate the fan adjusting parameter R; and when SW / S1 is smaller than or equal to 0.95, the electric control device calculates the fan adjusting parameter R = SS / S1, and a preset alarm short message is sent to a preset mobile phone number through a short message module. The invention further discloses a corresponding crop defrosting system and a crop defrosting method. According to the method, the frosting amount is judged without a deep learning algorithm, so that high energy consumption caused by a training process and deep learning is avoided; compared with deep learning, the method has the advantages that the calculated amount is reduced by several orders of magnitude, the cost is greatly reduced, the operation real-time performance is better, and the method is more suitable for real blowing defrosting control.

Description

Technical Field

[0001] The invention relates to the technical field of agricultural production, and in particular to a technology for preventing and controlling frost damage of crops. Background Art

[0002] When the temperature drops below 0℃, it may cause the water in plant cells to freeze, causing frost damage to crops. The impact of frost damage on crops is mainly manifested in cell dehydration, mechanical damage and wilting, and in severe cases, it can lead to the death of crops. Therefore, crop frost detection technology has been continuously developed in recent years. New crop frost detection technologies include: New technologies such as drone remote sensing detection technology, spectral cameras and thermal imaging camera technology have high implementation costs. They rely on image recognition and corresponding deep learning algorithms, have large amounts of computation and high energy consumption during implementation, and have a long interval from detection to response. Summary of the invention

[0003] The purpose of the present invention is to provide a method for detecting the amount of frost on a leaf surface, which can complete the task of fast and low-energy crop frost detection with a relatively low amount of calculation and provide a basis for the operation of a crop defrosting system.

[0004] The method for detecting the amount of frost on a leaf surface is characterized by being performed in the following steps: The first step is a preparation step, which includes an installation sub-step and a parameter determination sub-step; The installation sub-steps are: selecting a typical leaf from the canopy of a crop plant in the middle of the edge of the same crop plot in the farmland, setting a mounting frame on the ground at the location and fixing a grayscale camera on the mounting frame, wherein the grayscale camera has a flash; aiming the grayscale camera at the typical leaf and keeping the shooting angle and focal length fixed; selecting a suitable angle before fixing so that in the image captured by the grayscale camera, the brightness of the background other than the leaf is different from the brightness of the leaf; The grayscale camera is connected to an electronic control device; the electronic control device is connected to a text message module; The parameter determination sub-step is: photographing a typical leaf with the grayscale camera under frost-free conditions to obtain an 8-bit grayscale image of the typical leaf; manually analyzing the grayscale of each pixel in the image to determine that the grayscale value range of the leaf pixel under normal frost-free conditions is H1-H2.

[0005] The total number of pixels S1 in the leaf area of ​​the image is determined based on the grayscale values ​​of more than 99% of the background pixels falling outside H1-H2, and the grayscale values ​​of 99% of the leaves falling within H1-H2, and the grayscale values ​​of more than 99% of the background pixels being less than 230; wherein 0 The second step is a continuous detection step; the electronic control device controls the grayscale camera to take a grayscale image of a typical leaf every X minutes, 5 minutes ≤ X ≤ 60 minutes; after each shot, the electronic control device counts SW and SS;​ SW is the total number of pixels falling into the H1-H2 interval in the grayscale image at that time, and SW represents the number of unfrosted pixels on the leaf surface; SS is the number of pixels greater than or equal to 230 in the grayscale image, and SS represents the number of pixels in the frost layer in the image; When SW / S1>0.95, it means that the blade surface is basically frost-free, and the electronic control device does not calculate the fan adjustment parameter R at this time; When SW / S1≤0.95, the electronic control device calculates the fan adjustment parameter R=SS / S1, and sends a preset alarm SMS to a predetermined mobile phone number through the SMS module.

[0006] The present invention also aims to provide a crop defrosting system, which can quickly defrost crops after frost is detected, thereby alleviating frost damage and better protecting crops.

[0007] To achieve the above object, the present invention provides a crop defrosting system for implementing the above leaf frost amount detection method, wherein the electric control device is connected to a variable frequency fan, which is arranged above the farmland plot and whose blowing direction is inclined downward toward the crops; The air inlet of the variable frequency fan is selectively connected to the upper air inlet pipe and the lower air inlet pipe through a two-position three-way solenoid valve. The upper air inlet pipe is upwardly connected to an upper air inlet box, and the lower air inlet pipe is downwardly connected to a lower air inlet box. The outer walls of the upper air inlet box and the outer walls of the lower air inlet box are evenly provided with multiple air inlet holes. An upper temperature sensor is provided at the upper air inlet box, and a lower temperature sensor is provided at the lower air inlet box; The variable frequency fan, the upper temperature sensor, the lower temperature sensor and the two-position three-way solenoid valve are all connected with an electric control device.

[0008] The present invention also aims to provide a crop defrosting method, which coordinates frost amount detection and defrosting work to protect crops in time.

[0009] To achieve the above object, the crop defrosting method of the present invention adopts the above crop defrosting system, including a frequency control method and a ventilation control method; The frequency control method is: the maximum operating frequency of the variable frequency fan is MAX, and the minimum operating frequency of the variable frequency fan is MIN; the electronic control device controls the variable frequency fan to start, and the operating frequency is the higher value of R×MAX and MIN; when R>1, the electronic control device changes the value of R to 1; the variable frequency fan reduces frost damage by blowing wind to crops; During the startup of the variable frequency fan, take a grayscale image every 1 minute and calculate the SW / S1 value and R value. Correct the operating frequency of the variable frequency fan according to the new SW / S1 value and R value. When SW / S1>0.95, turn off the variable frequency fan and resume taking grayscale images every X minutes. The ventilation control method is: The temperature sensed by the upper temperature sensor is T1, and the temperature sensed by the lower temperature sensor is T2; When the variable frequency fan is turned on, the electronic control device compares T1 and T2. When T2≥T1 and T2>0℃, the electronic control device controls the two-position three-way solenoid valve to selectively connect the lower air inlet pipe; When T2<T1 and T1>0℃, the electronic control device controls the two-position three-way solenoid valve to selectively connect the upper air inlet duct.

[0010] The present invention has the following advantages: SS / S1 refers to the ratio of the number of pixels representing the frost layer to the number of pixels representing the leaf. The larger the ratio, the more severe the frost. The leaf surface frost amount detection method of the present invention can conveniently and quickly detect the R value, providing a basis for the operation and regulation of the crop defrosting system.

[0011] The present invention does not require a deep learning algorithm to determine the amount of frost, thus avoiding the high energy consumption caused by the training process and deep learning; the amount of calculation related to the frost amount performed by grayscale value is several orders of magnitude lower than that of deep learning, so that ordinary single-chip microcomputers are sufficient to implement the relevant algorithms in the present invention. Compared with the expensive industrial control computers required for deep learning, the cost is greatly reduced, and the real-time operation is better. From taking pictures to calculating the R value and the matching operating frequency of the variable frequency fan, this calculation process is real-time for people's normal perception, and does not require the long-term calculation required by deep learning, and is therefore more suitable for real blowing defrost control.

[0012] The upper air inlet box and the lower air inlet box prevent large impurities such as plant leaves from being sucked into the fan through the air inlet holes on their walls, ensuring the continuous and stable operation of the fan.

[0013] According to the size of the same crop plot in the farmland, multiple crop defrost systems are matched and set up to meet the overall needs of the crop plot to cope with frost damage.

[0014] The temperature at the surface is T1, and the temperature 10 meters above the surface is T2. When frost occurs, the probability of T2>T1 is higher, but there is still a possibility that T1>T2; the electronic control device compares T1 and T2, and chooses to draw in air with a higher temperature to blow towards the crops, which is more conducive to defrosting the crops.

[0015] By blowing wind over crops, frost damage can be reduced by either slowing down the rate at which frost forms on the leaves or reducing or even eliminating frost on the leaves.

[0016] The operating condition of the fan matches the current amount of frost. When there is more frost, the air volume is large, and when there is less frost, the air volume is small. This allows the air blowing defrost to automatically match the current frosting conditions, avoiding poor defrosting effect due to excessive frost and small air volume, and also avoiding waste of electricity due to small frost and large air volume.

[0017] During the startup of the variable frequency fan, a grayscale image is retaken every 1 minute, and a new frosting condition is obtained every 1 minute, so that the working state of the variable frequency fan can always match the current frost amount. After defrosting, grayscale images are taken every X minutes to reduce the working frequency of the grayscale camera. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a structural schematic diagram of the crop defrosting system in the present invention. Figure 1 The direction indicated by the middle arrow is the blowing direction of the variable frequency fan.

[0019] Figure 2 It is a schematic diagram of the electrical control structure of the crop defrosting system. DETAILED DESCRIPTION

[0020] like Figure 1 and Figure 2 As shown, the leaf surface frost amount detection method of the present invention is carried out in the following steps: The first step is a preparation step, which includes an installation sub-step and a parameter determination sub-step; The installation sub-steps are: select a typical leaf from the canopy of a crop plant in the middle of the edge of the same crop plot in the farmland, set a mounting frame 1 on the ground at that location and fix a grayscale camera 2 on the mounting frame 1, aim the grayscale camera 2 at the typical leaf and keep the shooting angle and focal length fixed; select a suitable angle before fixing so that in the image taken by the grayscale camera 2, the brightness of the background other than the leaf is different from the brightness of the leaf (preferably the difference in pixel grayscale value is greater than 10, and the difference in grayscale value of actual leaf pixels and background pixels is often greater than 100, especially under flash conditions); select leaves of moderate size, free of pests and diseases, uniform color, and easy to shoot as typical leaves.

[0021] The grayscale camera 2 is connected to the electronic control device 3; the grayscale camera 2 has a flash; the electronic control device 3 can be a single chip microcomputer or an industrial computer, preferably connected to a display screen. The electronic control device 3 is connected to the SMS module 4; the electronic control device 3 is preferably connected to a display screen 15, which can view various parameters.

[0022] The parameter determination sub-step is: photographing a typical leaf with the grayscale camera 2 (under the illumination of a flash) under the condition of no frost on the leaf surface, and obtaining an 8-bit grayscale image of the typical leaf; manually analyzing the grayscale of each pixel in the image, and determining that the grayscale value range of the leaf surface pixels under normal no frost conditions is H1-H2, including both end values, such as 50-150. Based on the actual range of the grayscale values ​​of the leaf surface pixels in the obtained grayscale image, the average grayscale value of the leaf surface pixels is P, and usually 99% of the grayscale values ​​of the leaf surface pixels fall within P±30.

[0023] When determining the values ​​of H1 and H2, the grayscale values ​​of more than 99% of the background pixels fall outside H1-H2 (excluding the two end values), and the grayscale values ​​of 99% of the leaves fall within H1-H2 (including the two end values), and the grayscale values ​​of more than 99% of the background pixels are less than 230 (above 230 is close to the white color of the frost layer), and the total number of pixels S1 in the leaf area in the image is determined; where 0 If the grayscale image taken by the grayscale camera 2 at a certain position, shooting angle and focal length does not meet the above standards (for example, there is a white background object or a background object close to the color of the leaves in the background), change the camera position and angle and shoot again until a suitable shooting position, shooting angle and shooting focal length are found, completing the parameter determination sub-step.

[0024] The flash can avoid abnormal grayscale values ​​of the leaf image caused by abnormal ambient brightness, and can also ensure that the grayscale value of the frost layer is greater than 230. According to experiments, under the condition of a flash, more than 99% of the grayscale values ​​of the pixels representing the frost layer are greater than or equal to 230. Typical leaves are selected at the edge of the same crop plot because when frost forms, the leaves at the edge of the crop plot usually frost earlier than the leaves in the middle of the plot. Typical leaves are selected in the middle of the edge of the same crop plot because the leaves in the middle of the edge of the plot are less susceptible to damage than the leaves at both ends of the edge of the plot. Grayscale camera 2 is selected because according to the method of the present invention, grayscale camera 2 can complete the task of detecting the amount of frost on the leaf surface, and has higher sensitivity and faster response speed than color cameras, and is cheaper (no filters and color processing circuits are required).

[0025] The second step is a continuous detection step; the electronic control device 3 controls the grayscale camera 2 to capture a grayscale image of a selected typical leaf every X minutes, 5 minutes ≤ X ≤ 60 minutes; after each capture, the electronic control device 3 counts SW and SS; SW is the total number of pixels falling into the H1-H2 interval in the grayscale image at that time, and SW represents the number of unfrosted pixels on the leaf surface; SS is the number of pixels greater than or equal to 230 in the grayscale image, and SS represents the number of pixels in the frost layer in the image; ​When SW / S1>0.95, it means that the blade surface is basically frost-free, and the electronic control device 3 does not calculate the fan adjustment parameter R at this time; When SW / S1≤0.95 (indicating that the blade surface may be frosted), the electronic control device 3 calculates the fan adjustment parameter R=SS / S1, and sends a preset alarm SMS to a predetermined mobile phone number through the SMS module 4.

[0026] SS / S1 refers to the ratio of the number of pixels representing the frost layer to the number of pixels representing the leaf. The larger the ratio, the more severe the frost. The leaf surface frost amount detection method of the present invention can conveniently and quickly detect the R value, providing a basis for the operation and regulation of the crop defrosting system.

[0027] The present invention does not require a deep learning algorithm to determine the amount of frost, thus avoiding the high energy consumption caused by the training process and deep learning; the amount of calculation related to the frost amount performed by grayscale value is several orders of magnitude lower than that of deep learning, so that ordinary single-chip microcomputers are sufficient to implement the relevant algorithms in the present invention. Compared with the expensive industrial control computers required for deep learning, the cost is greatly reduced, and the real-time operation is better. From taking pictures to calculating the R value and the matching operating frequency of the variable frequency fan, this calculation process is real-time for people's normal perception, and does not require the long-term calculation required by deep learning, and is therefore more suitable for real blowing defrost control.

[0028] The alarm text message can enable the staff to know in time whether the crops have been frosted.

[0029] The present invention also discloses a crop defrosting system for implementing the above-mentioned leaf surface frost amount detection method, wherein the electric control device 3 is connected to a variable frequency fan 5, and the variable frequency fan 5 is arranged above the farmland plot and its blowing direction is inclined downward toward the crop 6; The air inlet of the variable frequency fan 5 is selectively connected to the upper air inlet pipe 8 and the lower air inlet pipe 9 through a two-position three-way solenoid valve 7. The upper air inlet pipe 8 is upwardly connected to an upper air inlet box 10, and the lower air inlet pipe 9 is downwardly connected to a lower air inlet box 11. The outer walls of the upper air inlet box 10 and the outer walls of the lower air inlet box 11 are evenly provided with a plurality of air inlet holes 12. An upper temperature sensor 13 is provided at the upper air inlet box 10, and a lower temperature sensor 14 is provided at the lower air inlet box 11; The variable frequency fan 5, the upper temperature sensor 13, the lower temperature sensor 14 and the two-position three-way solenoid valve 7 are all connected to an electric control device 3; The temperature sensed by the upper temperature sensor 13 is T1, and the temperature sensed by the lower temperature sensor 14 is T2; During defrosting, the electronic control device 3 compares T1 and T2. When T2≥T1 and T2>0℃, the electric control device 3 controls the two-position three-way solenoid valve 7 to selectively connect the lower air inlet pipe 9; When T2<T1 and T1>0℃, the electric control device 3 controls the two-position three-way solenoid valve 7 to selectively connect the lower air inlet pipe 9; The variable frequency fan 5, the upper air inlet box 10 and the lower air inlet box 11 are all supported by a fixing frame; the fixing frame is conventional technology and is not shown in the figure.

[0030] According to the size of the same crop 6 plot in the farmland, multiple crop 6 defrost systems are matched and set up, and the overall need of coping with frost damage of the crop 6 plot is matched through the multiple crop 6 defrost systems.

[0031] The upper air inlet box 10 and the lower air inlet box 11 prevent large impurities such as plant leaves from being sucked into the fan through the air inlet holes 12 on the walls thereof, thereby ensuring that the fan works continuously and stably.

[0032] The present invention also provides a crop defrosting method, which is performed by using the above crop defrosting system according to the following steps: including a frequency control method and a ventilation control method; (I) The frequency control method is: the maximum operating frequency of the variable frequency fan 5 is MAX, and the minimum operating frequency of the variable frequency fan 5 is MIN; the electronic control device 3 controls the variable frequency fan 5 to start, and the operating frequency is the higher value of R×MAX and MIN; when R>1, the electronic control device 3 modifies the value of R to 1. R>1 means that the environment of the blade, such as the ground surface, is also frosted, and the reflection of the frosted part enters the lens of the grayscale camera; Although the ground or background is frosted, if the reflection of the flash does not enter the lens of the grayscale camera, the grayscale value of the pixel corresponding to the frosted part of the ground or background is usually much lower than 230. The grayscale camera is aimed at a typical leaf, so there is no problem that the reflection of the typical leaf cannot enter the lens.

[0033] The variable frequency fan 5 reduces frost damage by blowing air to the crops 6; During the startup of the variable frequency fan 5, a grayscale image is re-taken every 1 minute and the SW / S1 value and R value are calculated. The operating frequency of the variable frequency fan 5 is corrected according to the new SW / S1 value and R value. When SW / S1>0.95, the variable frequency fan 5 is turned off and the grayscale image is taken every X minutes again. (II) The ventilation control method is: The temperature sensed by the upper temperature sensor 13 is T1, and the temperature sensed by the lower temperature sensor 14 is T2; When the variable frequency fan 5 is turned on, the electronic control device 3 compares T1 and T2. When T2≥T1 and T2>0℃, the electric control device 3 controls the two-position three-way solenoid valve 7 to selectively connect the lower air inlet pipe 9; When T2<T1 and T1>0°C, the electronic control device 3 controls the two-position three-way solenoid valve 7 to selectively connect the upper air inlet pipe 8.

[0034] The temperature at the surface is T1, and the temperature 10 meters above the surface is T2. When frost occurs, the probability of T2>T1 is high, but there is still a possibility that T1>T2; the electronic control device 3 compares T1 and T2, and chooses to draw in air with a higher temperature to blow toward the crops 6, which is more conducive to defrosting the crops 6.

[0035] By blowing wind toward the crops 6, the speed of frost formation on the leaves is reduced, or the frost on the leaves is reduced or even eliminated, thereby reducing frost damage.

[0036] The operating condition of the fan matches the current amount of frost. When there is more frost, the air volume is large, and when there is less frost, the air volume is small. This allows the air blowing defrost to automatically match the current frosting conditions, avoiding poor defrosting effect due to excessive frost and small air volume, and also avoiding waste of electricity due to small frost and large air volume.

[0037] During the startup of the variable frequency fan 5, a grayscale image is retaken every 1 minute, and a new frosting condition is acquired every 1 minute, so that the working state of the variable frequency fan 5 can always match the current frost amount. After defrosting, grayscale images are taken every X minutes to reduce the working frequency of the grayscale camera 2.

[0038] The above embodiments are only used to illustrate rather than limit the technical solutions of the present invention. Although the present invention is described in detail with reference to the above embodiments, those skilled in the art should understand that the present invention can still be modified or replaced by equivalents. Any modification or partial replacement that does not depart from the spirit and scope of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for detecting the amount of frost on a leaf surface, characterized in that Follow these steps: The first step is a preparation step, which includes an installation sub-step and a parameter determination sub-step; The installation sub-step is: select a typical leaf from the canopy of a crop plant in the middle of the edge of the same crop plot in the farmland, set a mounting frame on the ground at the location and fix the grayscale camera on the mounting frame, aim the grayscale camera at the typical leaf and keep the shooting angle and focal length fixed; before fixing, select a suitable angle so that in the image captured by the grayscale camera, the brightness of the background other than the leaf is different from the brightness of the leaf; The grayscale camera is connected to an electronic control device; the grayscale camera has a flash light; the electronic control device is connected to a text message module; The parameter determination sub-step is: photographing a typical leaf with the grayscale camera under the condition that the leaf surface is frost-free to obtain an 8-bit grayscale image of the typical leaf; Manually analyze the grayscale of each pixel in the image, determine that the grayscale value range of leaf surface pixels under normal frost-free conditions is H1-H2, and that more than 99% of the grayscale values ​​of the background pixels fall outside H1-H2, and 99% of the grayscale values ​​of the leaves fall within H1-H2, and more than 99% of the grayscale values ​​of the background pixels are less than 230, and determine the total number of pixels S1 in the leaf area in the image; where 0<H1<H2<255; The second step is a continuous detection step; the electronic control device controls the grayscale camera to take a grayscale image of a typical leaf every X minutes, 5 minutes ≤ X ≤ 60 minutes; after each shot, the electronic control device counts SW and SS; SW is the total number of pixels falling into the H1-H2 interval in the grayscale image at that time, and SW represents the number of unfrosted pixels on the leaf surface; SS is the number of pixels greater than or equal to 230 in the grayscale image, and SS represents the number of pixels in the frost layer in the image; When SW / S1>0.95, it means that the blade surface is basically frost-free, and the electronic control device does not calculate the fan adjustment parameter R at this time; When SW / S1≤0.95, the electronic control device calculates the fan adjustment parameter R=SS / S1, and sends a preset alarm SMS to a predetermined mobile phone number through the SMS module.

2. A crop defrosting system, used to implement the above-mentioned leaf surface frost amount detection method, wherein the electronic control device is connected to a variable frequency fan, which is arranged above the farmland plot and whose blowing direction is inclined downward toward the crops; Features: The air inlet of the variable frequency fan is selectively connected to the upper air inlet pipe and the lower air inlet pipe through a two-position three-way solenoid valve. The upper air inlet pipe is upwardly connected to an upper air inlet box, and the lower air inlet pipe is downwardly connected to a lower air inlet box. The outer walls of the upper air inlet box and the outer walls of the lower air inlet box are evenly provided with multiple air inlet holes. An upper temperature sensor is provided at the upper air inlet box, and a lower temperature sensor is provided at the lower air inlet box; The variable frequency fan, the upper temperature sensor, the lower temperature sensor and the two-position three-way solenoid valve are all connected with an electric control device.

3. A crop defrosting method, using the crop defrosting system as claimed in claim 2, characterized in that: Including frequency control method and ventilation control method; The frequency control method is: the maximum operating frequency of the variable frequency fan is MAX, and the minimum operating frequency of the variable frequency fan is MIN; the electronic control device controls the variable frequency fan to start, and the operating frequency is the higher value of R×MAX and MIN; when R>1, the electronic control device changes the value of R to 1; the variable frequency fan reduces frost damage by blowing wind to crops; During the startup of the variable frequency fan, take a grayscale image every 1 minute and calculate the SW / S1 value and R value. Correct the operating frequency of the variable frequency fan according to the new SW / S1 value and R value. When SW / S1>0.95, turn off the variable frequency fan and resume taking grayscale images every X minutes. The ventilation control method is: The temperature sensed by the upper temperature sensor is T1, and the temperature sensed by the lower temperature sensor is T2; When the variable frequency fan is turned on, the electronic control device compares T1 and T2. When T2≥T1 and T2>0℃, the electronic control device controls the two-position three-way solenoid valve to selectively connect the lower air inlet pipe; When T2<T1 and T1>0℃, the electronic control device controls the two-position three-way solenoid valve to selectively connect the upper air inlet duct.

Images