Water pump control system and method based on intelligent farmland management

Abstract

The invention relates to a water pump control system and method based on intelligent farmland management, and the system comprises a camera which is used for collecting an image of rice in a farmland, and obtaining a rice image; the liquid level sensor is used for collecting water depth information of a farmland to obtain water depth data; the humidity sensor is used for collecting the humidity of farmland soil to obtain humidity data; the local server is used for identifying the growth cycle of the rice according to the rice image and generating water supply demand information according to the growth cycle of the rice, the water depth data and the humidity data; the cloud server is used for generating a water supply control instruction according to the water supply demand information; the controller is in communication connection with the cloud server and used for controlling starting and stopping of the water pump according to the water supply control instruction so as to control the water supply amount of the water pump; the technical problems that water supply of a water pump needs to be controlled manually according to real-time data for rice field water supply, and unmanned and intelligent requirements required by intelligent farmland management cannot be met are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent agriculture, and particularly relates to a water pump control system and method based on intelligent farmland management. Background Art

[0002] Existing intelligent farmlands monitor various parameters of the farmland and various parameters of the rice in the farmland, and upload the parameters to the Internet in real time, facilitating farmland managers to understand the growth status of the rice in real time through intelligent communication terminals.

[0003] However, in the prior art, the water supply of the paddy field also requires manual control of the water pump according to real-time data, which cannot meet the requirements of unmanned and intelligent management of smart farmlands. Summary of the Invention

[0004] In order to solve the technical problems in the prior art that the water supply of the paddy field requires manual control of the water pump according to real-time data and cannot meet the requirements of unmanned and intelligent management of smart farmlands, the present invention provides a water pump control system and method based on intelligent farmland management.

[0005] The technical solution of the present invention for solving the above technical problems is as follows:

[0006] A water pump control system based on intelligent farmland management, comprising:

[0007] A camera, configured to collect images of the rice in the farmland to obtain rice images;

[0008] A liquid level sensor, configured to collect water depth information of the farmland to obtain water depth data;

[0009] A humidity sensor, configured to collect the humidity of the soil in the farmland to obtain humidity data;

[0010] A local server, connected to the camera, the liquid level sensor, and the humidity sensor, configured to identify the growth cycle of the rice according to the rice images, and generate a water supply demand information according to the growth cycle of the rice, the water depth data, and the humidity data;

[0011] A cloud server, establishing a communication connection with the local server, configured to generate a water supply control instruction according to the water supply demand information;

[0012] A controller, establishing a communication connection with the cloud server, configured to control the start and stop of the water pump according to the water supply control instruction to control the water supply volume of the water pump.

[0013] The beneficial effects of the present invention are as follows: By using a computer vision recognition model to recognize the growth cycle of rice and controlling the start and stop of the water pump according to the growth cycle of rice to control the water supply of the water pump, automatic water supply on demand at each growth stage of rice is realized; it solves the technical problems that the water supply of paddy fields requires manual control of the water pump according to real-time data and cannot meet the requirements of unmanned and intelligent management of smart farms.

[0014] Based on the above technical solution, the present invention can be further improved as follows.

[0015] Further, the local server is specifically used to construct a rice height recognition model, a rice grain recognition model, and a rice maturity recognition model; use the rice height recognition model to recognize the growth height of rice; use the rice grain recognition model to recognize whether there are rice grains on the rice, and obtain a rice grain recognition result; use the rice maturity recognition model to recognize the maturity of the rice grains on the rice, and obtain a maturity recognition result; judge the growth cycle of the rice according to the rice growth height, the rice grain recognition result, and the maturity recognition result.

[0016] Further, the local server is specifically used to construct a first computer vision recognition model; collect rice images of the rice at different growth cycles to obtain a plurality of height training data sets; wherein, each height data training set corresponds to a growth cycle of the rice; use the plurality of height training data sets to train the first computer vision recognition model to obtain the rice height recognition model.

[0017] Further, the local server is specifically used to construct a second computer vision recognition model; collect rice images of the rice during the filling stage and the maturity stage to obtain a rice grain training data set; use the rice grain training data set to train the second computer vision recognition model to obtain the rice grain recognition model.

[0018] Further, the local server is specifically used to construct a second computer vision recognition model; collect rice images of the rice at the maturity stage to obtain a rice maturity training data set; use the rice maturity training data set to train the second computer vision recognition model to obtain the rice maturity recognition model.

[0019] Further, the local server is specifically used to judge whether it is in the stem growth stage according to the rice growth height. If so, it is determined that the growth cycle of the rice is in the stem growth stage; if not, it is judged whether it is in the filling stage according to the rice grain recognition result and the maturity recognition result. If so, it is determined that the rice is in the filling stage; if not, it is determined that the rice is in the maturity stage.

[0020] Further, when it is determined that the growth cycle of the rice is in the stem growth stage, the first water supply demand information is generated according to the water depth data; when it is determined that the growth cycle of the rice is in the filling stage, the second water supply demand information is generated according to the water depth data; when it is determined that the growth cycle of the rice is in the maturity stage, the third water supply demand information is generated according to the humidity data; wherein, the water supply demand information includes the first water supply demand information, the second water supply demand information, and the third water supply demand information.

[0021] Further, it further includes:

[0022] A mobile communication terminal, which establishes a communication connection with the cloud server and is used to send a water pump control request to the cloud server;

[0023] The cloud server is further used to obtain the water pump control request through the wireless communication terminal and generate the water supply control instruction according to the water pump control request.

[0024] To solve the above technical problems, the present invention also provides a water pump control method applied to the above water pump control system based on intelligent farmland management, and its specific technical content is as follows:

[0025] A water pump control method applied to the above water pump control system based on intelligent farmland management, including the following steps:

[0026] Collect an image of the rice in the farmland to obtain a rice image;

[0027] Collect the water depth information of the farmland to obtain water depth data;

[0028] Collect the humidity of the soil in the farmland to obtain humidity data;

[0029] Identify the growth cycle of the rice according to the rice image, and generate water supply demand information according to the growth cycle of the rice, the water depth data, and the humidity data;

[0030] Generate a water supply control instruction according to the water supply demand information;

[0031] Control the start and stop of the water pump according to the water supply control instruction.

[0032] To solve the above technical problems, the present invention also provides a smart farmland management system, and its specific technical content is as follows:

[0033] A smart farmland management system includes multiple subsystems, and at least one of the systems is the above water pump control system based on intelligent farmland management. Description of the Drawings

[0034] Figure 1This is the structural schematic diagram of a water pump control system based on intelligent farmland management in an embodiment of the present invention. Detailed implementation manners

[0035] The principles and features of the present invention will be described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not intended to limit the scope of the present invention.

[0036] As Figure 1 shown, this embodiment provides a water pump control system based on intelligent farmland management, including:

[0037] A camera for collecting images of rice in the farmland to obtain rice images;

[0038] A liquid level sensor for collecting water depth information of the farmland to obtain water depth data;

[0039] A humidity sensor for collecting the humidity of the soil in the farmland to obtain humidity data;

[0040] A local server connected to the camera, the liquid level sensor, and the humidity sensor, for identifying the growth cycle of the rice according to the rice images, and generating water supply demand information according to the growth cycle of the rice, the water depth data, and the humidity data;

[0041] A cloud server establishing a communication connection with the local server, for generating a water supply control instruction according to the water supply demand information;

[0042] A controller establishing a communication connection with the cloud server, for controlling the start and stop of the water pump according to the water supply control instruction to control the water supply volume of the water pump.

[0043] By using a computer vision recognition model to identify the growth cycle of rice, and controlling the start and stop of the water pump according to the growth cycle of rice to control the water supply volume of the water pump, it realizes automatic water supply on demand at each growth stage of rice; solves the technical problems that the water supply of paddy fields needs to be manually controlled by real-time data, and cannot meet the requirements of unmanned and intelligent management of smart farmland.

[0044] In some embodiments, the local server is specifically configured to construct a rice height recognition model, a rice grain recognition model, and a rice maturity recognition model; use the rice height recognition model to recognize the growth height of the rice; use the rice grain recognition model to recognize whether there are rice grains on the rice, and obtain a rice grain recognition result; use the rice maturity recognition model to recognize the maturity of the rice grains on the rice, and obtain a maturity recognition result; and judge the growth cycle of the rice according to the rice growth height, the rice grain recognition result, and the maturity recognition result. The rice height recognition model, the rice grain recognition model, and the rice maturity recognition model all adopt a convolutional neural network model.

[0045] The specific method for model processing is as follows: (1) Use image subtraction and color clustering to remove the background from the rice image to be recognized; (2) Perform image segmentation on the rice image after background removal; (3) Extract the edges of the segmented image, and color the edge region using the four-color method; (4) Identify the growth characteristics of the rice seedlings in the four color domains, and judge the stage in the growth period of the rice. When using the rice grain recognition model to recognize the growth cycle of the rice, it is necessary to recognize the shape of the rice grains in the image. When a graphic with the shape of rice grains is recognized in the image, it is determined that rice has grown on the rice, and it is judged that the rice is in the filling stage. Use the rice maturity recognition model to recognize the color of the rice grains on the rice to determine whether the rice grains are mature. When it is determined that the color of the rice grains is golden yellow, it is determined that the rice is in the mature stage.

[0046] In some embodiments, the local server is specifically configured to construct a first computer vision recognition model; collect rice images of the rice at different growth cycles to obtain a plurality of height training data sets; wherein, each of the height data training sets corresponds to a growth cycle of the rice; and use the plurality of height training data sets to train the first computer vision recognition model to obtain the rice height recognition model.

[0047] Collect rice images of the rice at different growth cycles to obtain a plurality of height training data sets; specifically:

[0048] Collect rice images of the rice in the seedling stage to obtain a first height training data set;

[0049] Collect rice images of the rice in the tillering stage to obtain a second height training data set;

[0050] Collect rice images of the rice in the jointing stage to obtain a third height training data set;

[0051] Collect rice images of the rice in the booting stage to obtain a fourth height training data set;

[0052] Collect rice images of the rice in the heading stage to obtain a fifth height training data set;

[0053] Collect rice images of the rice at the flowering stage to obtain the sixth height training dataset. The multiple height training datasets are respectively the first height training dataset, the second height training dataset, the third height training dataset, the fourth height training dataset, the fifth height training dataset, and the sixth height training dataset.

[0054] In some embodiments, the local server is specifically configured to construct a second computer vision recognition model; collect rice images of the rice at the filling stage and the maturity stage to obtain a rice grain training dataset; and use the rice grain training dataset to train the second computer vision recognition model to obtain the rice grain recognition model.

[0055] In some embodiments, the local server is specifically configured to construct a second computer vision recognition model; collect rice images of the rice at the maturity stage to obtain a rice maturity training dataset; and use the rice maturity training dataset to train the second computer vision recognition model to obtain the rice maturity recognition model.

[0056] In some embodiments, the local server is specifically configured to determine whether it is in the stem growth stage according to the rice growth height. If so, it is determined that the growth cycle of the rice is in the stem growth stage; if not, it is determined whether it is in the filling stage according to the rice grain recognition result and the maturity recognition result. If so, it is determined that the rice is in the filling stage; if not, it is determined that the rice is in the maturity stage. The local server is specifically configured to determine that the growth cycle of the rice is in the stem growth stage when the rice growth height is less than or equal to a preset height. The stem growth stage includes the seedling stage, the tillering stage, the jointing stage, the booting stage, the heading stage, and the flowering stage.

[0057] In some embodiments, when it is determined that the growth cycle of the rice is in the stem growth stage, the first water supply demand information is generated according to the water depth data; when it is determined that the growth cycle of the rice is in the filling stage, the second water supply demand information is generated according to the water depth data; when it is determined that the growth cycle of the rice is in the maturity stage, the third water supply demand information is generated according to the humidity data; wherein, the water supply demand information includes the first water supply demand information, the second water supply demand information, and the third water supply demand information. When the rice is in the stem growth stage, the water depth of the paddy field is kept at 8 cm - 10 cm deep. When the water depth collected by the liquid level sensor exceeds the range of 10 cm, the controller controls the water pump to pump out the water in the paddy field according to the first water supply demand information so that the water depth of the paddy field is kept at 8 cm - 10 cm; when the water depth of the paddy field is lower than 8 cm, the controller controls the water pump to supply water to the paddy field according to the first water supply demand information so that the water depth of the paddy field is kept at 8 cm - 10 cm.

[0058] When the rice is in the filling stage, the water depth in the paddy field is maintained at 3 cm - 5 cm. When the water depth collected by the liquid level sensor exceeds the range of 5 cm, the controller controls the water pump to pump out the water in the paddy field according to the second water supply demand information so that the water depth in the paddy field is maintained at 3 cm - 5 cm; when the water depth in the paddy field is lower than 3 cm, the controller controls the water pump to supply water to the paddy field according to the second water supply demand information so that the water depth in the paddy field is maintained at 3 cm - 5 cm.

[0059] When the rice is in the mature stage, the water depth in the paddy field is maintained at 0 cm, that is, the paddy field does not need to store water, and only the soil humidity of the paddy field needs to be maintained. When the humidity of the soil in the paddy field collected by the humidity sensor is lower than the preset lower humidity limit value, the controller controls the water pump to work to supply water to the paddy field through the water pump, so as to increase the humidity of the soil in the paddy field.

[0060] In some embodiments, it further includes:

[0061] A mobile communication terminal, which establishes a communication connection with the cloud server and is used to send a water pump control request to the cloud server;

[0062] The cloud server is further used to obtain the water pump control request through the wireless communication terminal and generate the water supply control instruction according to the water pump control request.

[0063] The control system further includes a temperature sensor. The temperature sensor collects the air temperature near the paddy field to obtain temperature data. The local server also receives the rice image, water depth data, humidity data and temperature data and transmits them to the cloud server through the Internet. The mobile terminal receives the rice image, water depth data, humidity data and temperature data by interacting with the cloud server; and displays the rice image, water depth data, humidity data and temperature data on the mobile terminal.

[0064] In some other embodiments, a water pump control method applied to the above water pump control system based on intelligent farmland management is also disclosed, including the following steps:

[0065] Collect an image of the rice in the farmland to obtain a rice image;

[0066] Collect the water depth information of the farmland to obtain water depth data;

[0067] Collect the humidity of the soil in the farmland to obtain humidity data;

[0068] Identify the growth cycle of the rice according to the rice image, and generate water supply demand information according to the growth cycle of the rice, the water depth data and the humidity data;

[0069] Generate a water supply control instruction according to the water supply demand information;

[0070] Control the start and stop of the water pump according to the water supply control instruction.

[0071] In some other embodiments, an intelligent farmland management system includes multiple subsystems, and at least one of the systems is the water pump control system based on intelligent farmland management described above.

[0072] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the concept and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A water pump control system based on intelligent farmland management, characterized in that, Including: A camera for collecting images of rice in farmland to obtain rice images; A liquid level sensor for collecting water depth information of farmland to obtain water depth data; A humidity sensor for collecting the humidity of the soil in farmland to obtain humidity data; A local server connected to the camera, the liquid level sensor and the humidity sensor, for identifying the growth cycle of the rice according to the rice images, and generating a water supply demand information according to the growth cycle of the rice, the water depth data and the humidity data; A cloud server establishing a communication connection with the local server, for generating a water supply control instruction according to the water supply demand information; A controller establishing a communication connection with the cloud server, for controlling the start and stop of a water pump according to the water supply control instruction to control the water supply volume of the water pump.

2. The water pump control system based on intelligent farmland management according to claim 1, wherein The local server is specifically used for constructing a rice height recognition model, a rice grain recognition model and a rice maturity recognition model; using the rice height recognition model to identify the growth height of rice; using the rice grain recognition model to identify whether there are rice grains on the rice to obtain a rice grain recognition result; using the rice maturity recognition model to identify the maturity of the rice grains on the rice to obtain a maturity recognition result; judging the growth cycle of the rice according to the rice growth height, the rice grain recognition result and the maturity recognition result.

3. The water pump control system based on intelligent farmland management according to claim 2, characterized in that The local server is specifically used for constructing a first computer vision recognition model; collecting rice images of the rice in different growth cycles to obtain a plurality of height training data sets; wherein, each of the height data training sets corresponds to a growth cycle of the rice; using the plurality of height training data sets to train the first computer vision recognition model to obtain the rice height recognition model.

4. The water pump control system based on intelligent farmland management according to claim 2, wherein The local server is specifically used for constructing a second computer vision recognition model; collecting rice images of the rice in the filling stage and the maturity stage to obtain a rice grain training data set; using the rice grain training data set to train the second computer vision recognition model to obtain the rice grain recognition model.

5. The water pump control system based on intelligent farmland management according to claim 2, wherein The local server is specifically used for constructing a second computer vision recognition model; collecting rice images of the rice in the maturity stage to obtain a rice maturity training data set; using the rice maturity training data set to train the second computer vision recognition model to obtain the rice maturity recognition model.

6. The water pump control system based on intelligent farmland management according to claim 2, wherein The local server is specifically used for judging whether it is in the stem growth stage according to the rice growth height; if so, determining that the growth cycle of the rice is in the stem growth stage; if not, judging whether it is in the filling stage according to the rice grain recognition result and the maturity recognition result; if so, determining that the rice is in the filling stage; if not, determining that the rice is in the maturity stage.

7. The water pump control system based on intelligent farmland management according to claim 6, wherein When it is determined that the growth cycle of the rice is in the stem growth stage, the first water supply demand information is generated according to the water depth data; when it is determined that the growth cycle of the rice is in the filling stage, the second water supply demand information is generated according to the water depth data; when it is determined that the growth cycle of the rice is in the mature stage, the third water supply demand information is generated according to the humidity data; wherein, the water supply demand information includes the first water supply demand information, the second water supply demand information and the third water supply demand information.

8. The water pump control system based on intelligent farmland management according to claim 1, wherein It further includes: A mobile communication terminal, which establishes a communication connection with the cloud server and is used to send a water pump control request to the cloud server; The cloud server is further used to obtain the water pump control request through the wireless communication terminal and generate the water supply control instruction according to the water pump control request.

9. A water pump control method applied to the water pump control system based on intelligent farmland management according to any one of claims 1 to 8, characterized in that, It includes the following steps: Collect the image of the rice in the farmland to obtain the rice image; Collect the water depth information of the farmland to obtain the water depth data; Collect the humidity of the soil in the farmland to obtain the humidity data; Identify the growth cycle of the rice according to the rice image, and generate water supply demand information according to the growth cycle of the rice, the water depth data and the humidity data; Generate a water supply control instruction according to the water supply demand information; Control the start and stop of the water pump according to the water supply control instruction to control the water supply volume of the water pump.

10. A smart farmland management system, characterized in that, It includes a plurality of subsystems, and at least one of the systems is the water pump control system based on intelligent farmland management according to any one of claims 1 to 8.

Images