Information processing method and system based on Internet of Things
By identifying the crop species and growth status in the farmland image, and combining weather and soil data, the harvesting suitability is calculated, which solves the problem of insufficient optimization of weather factors in the existing technology, and improves the accuracy of harvesting time and farmland Management efficiency.
Patent Information
- Application Number
- CN202411140252.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-11-29
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing IoT-based agricultural harvest monitoring methods have failed to effectively optimize the degree of crop damage in combination with weather factors, resulting in a decrease in the accuracy of harvest time.
By obtaining farmland images, the crop growth state acquisition model is used to identify crop species and growth state, and combining the weather data and soil moisture content and degree of lodging within the harvest time interval, the harvest suitability is calculated to determine the optimal harvest time.
It improves the accuracy of crop harvesting time and farmland management efficiency, ensures the optimal harvesting conditions for crops, and reduces the impact of harvesting difficulty and yield due to weather changes.
Smart Images

Figure CN119048266A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of crop harvesting information processing, and in particular to an information processing method and system based on the Internet of Things. Background Art
[0002] The Internet of Things refers to various physical devices and objects that are connected and communicate with each other through the Internet. These devices can be smart devices, sensors, machines, etc. IoT technology enables these devices to collect and exchange data and realize the interconnection of information to achieve smarter and more automated functions and services.
[0003] As IoT technology matures, it has also been applied to the agricultural field, including farmland irrigation, farmland pesticide application and farmland harvesting, bringing many innovations and improvements to modern agriculture. For farmland harvesting, IoT can monitor the growth status of crops based on the collected data and determine whether harvesting is needed.
[0004] At present, there are documents on agricultural harvest monitoring based on the Internet of Things. For example, the Chinese patent with publication number CN116109268B proposes a smart agricultural supervision system and method based on the Internet of Things. This method analyzes various information and image data during the growth of crops; and analyzes the maturity of crop growth based on the analysis results to provide data support for the harvesting of later crops. Another example is the Chinese patent with publication number CN110514248B, which proposes an intelligent and automated agricultural data collection system based on big data. This system collects different data in the environment and calculates the impact of different environmental data on the growth cycle of crops, so that the growth cycle of crops under the current environmental factors can be accurately calculated, and then the appropriate harvesting time can be calculated. The above documents only monitor the growth status of crops to judge the appropriate harvesting time, but ignore the impact of weather on harvesting in actual harvesting, which will reduce the accuracy of crop harvesting. In addition, the literature ("Cross-regional operation scheduling of combine harvesters under the influence of weather changes", "Nanjing Agricultural University", Peng Shuqin, 2022-05-01) considers the impact of weather changes on the cross-regional operation scheduling of harvesters, establishes a scheduling model for cross-regional operations of combine harvesters under rain and other severe weather conditions, and combines simulation analysis to obtain the optimal scheduling solution. The above method only schedules the operation of harvesters based on weather changes, but does not optimize the harvesting time based on the degree of damage to crops caused by weather factors, which will reduce the accuracy of crop harvesting.
[0005] To this end, an information processing method and system based on the Internet of Things are proposed. Summary of the invention
[0006] The purpose of the present invention is to provide an information processing method and system based on the Internet of Things. First, a farmland image is obtained, and the type of crops in the farmland and the current growth status of the crops are obtained according to the farmland image; secondly, when the current growth status of the crops is mature, the harvesting time interval corresponding to the type of crops in the farmland is obtained. dq ,time dq +T zl ] within the harvesting environment data; secondly, obtaining the harvesting suitability according to various weather factors in the harvesting environment data and the soil moisture content and lodging degree of the farmland itself; finally, sorting the harvesting suitability from high to low, displaying the sorted harvesting suitability and the corresponding harvesting time, and displaying the suitable harvesting time.
[0007] To achieve the above object, the present invention provides the following technical solutions:
[0008] An information processing method based on the Internet of Things, comprising:
[0009] Acquire farmland images, and acquire the types of crops in the farmland and the current growth status of the crops based on the farmland images.
[0010] Furthermore, the specific steps include: obtaining a farmland image taken by a drone; inputting the farmland image into a crop growth state acquisition model to obtain the crop type and current growth state; the crop growth state acquisition model includes an input layer, a crop detection layer, a crop feature extraction layer, a crop feature fusion layer, a crop type acquisition layer, a crop growth state acquisition layer and an output layer; the input layer is used to input the farmland image into the crop growth state acquisition model; the crop detection layer is used to obtain N crop images based on the farmland image;
[0011] The crop feature extraction layer is used to obtain crop feature maps of N crop images based on the crop images; the crop feature fusion layer is used to perform feature fusion on the crop feature maps corresponding to each crop image to obtain crop features corresponding to the N crop images; the crop features include crop color features, crop leaf texture features, crop leaf shape features and crop leaf size features;
[0012] The crop type acquisition layer is used to acquire the crop types corresponding to the N crop images according to the crop characteristics, and acquire the farmland crop types corresponding to the farmland image;
[0013] The crop growth status acquisition layer is used to acquire the current growth status of the crops according to the crop types in the farmland and the crop characteristics; the output layer is used to output the crop types in the farmland and the current growth status.
[0014] Furthermore, the crop type acquisition layer further comprises: when the crop types corresponding to the N crop images are different, constructing a type set (ZL1, ZL2, ..., ZL M ), where M represents the number of crop types identified by the N crop images, and ZL M It is represented by the number of the Mth crop species identified in the N crop images;
[0015] Get the maximum value ZL in the category set max ; When ZL max When / N is greater than the set category ratio threshold, the ZL max The corresponding crop type is used as the crop type of the farmland; the ZL max The crop features of the corresponding crop images are input into the crop growth status acquisition layer;
[0016] When ZL max When / N is not greater than the set category proportion threshold, the crop detection layer reacquires N crop images.
[0017] Furthermore, when the number of times the crop detection layer reacquires N crop images reaches a set repetition threshold, the crop detection layer directly jumps to the output layer to output an abnormal crop type warning.
[0018] Furthermore, the crop growth status acquisition layer further includes: comparing each of the crop features received by the crop growth status acquisition layer with the crop maturity features corresponding to the crop types in the farmland to obtain feature similarity; when the feature similarity is greater than a set similarity threshold, it is indicated that the growth state corresponding to the crop feature is mature; obtaining the number of the crop features with a mature growth state, recorded as sl cs ; When sl cs / sl js When the ratio of the crop to the mature state is greater than the set maturity state ratio threshold, the current growth state of the crop is mature. js It represents the number of the crop features received by the crop growth state acquisition layer; otherwise, the current growth state of the crop is immature.
[0019] When the current growth state of the crop is mature, obtain the harvest time interval corresponding to the crop type in the farmland [time dq ,time dq +T zl ] within the harvesting environment data.
[0020] Furthermore, timedq Indicates the current time; T zl It is represented by the maximum harvesting time interval corresponding to the crop type in the farmland, T zl Is a positive integer.
[0021] Furthermore, the harvesting environment data includes the current farmland soil moisture content, the current crop lodging area, the daily rainfall within the harvesting time interval, the average temperature, the light time, the light intensity and the maximum wind speed.
[0022] The harvesting suitability is obtained according to the harvesting environment data.
[0023] Furthermore, the harvesting suitability of each day within the harvesting time interval is obtained according to the harvesting environment, and the calculation formula of the harvesting suitability is:
[0024]
[0025] in, It is expressed as the harvest suitability of the i-th day after the current time, and the value of i ranges from 0 to T zl ; When the value of i is 0, Represents the harvest suitability at the current time; trhs i 、jyl i 、pjwd i 、fs i and dfmj i They respectively represent the soil moisture content, rainfall, average temperature, maximum wind speed and lodging degree of the i-th day after the current time; bzwd represents the standard maturity temperature value corresponding to the set type of crop in the farmland; α1, α2, α3, α4 and α5 represent the first soil moisture coefficient, the first rainfall coefficient, the first temperature coefficient, the first wind speed coefficient and the first lodging degree coefficient respectively; exp() represents an exponential function with natural constants as the base.
[0026] Furthermore, the soil moisture content trhs i Including: When i = 0, trhs i It is expressed as the soil moisture content at the current time; when i>0, trhs i The calculation formula is:
[0027]
[0028] Among them, jyl i 、pjwd i 、gzsj i and gzqd i They represent the rainfall, average temperature, sunshine time and sunshine intensity of the ith day after the current time; trhsi-1 It is represented as the soil moisture content on the i-1th day after the current time; β1, β2, β3 and β4 represent the second soil moisture coefficient, the second rainfall coefficient, the second average temperature coefficient and the light coefficient respectively; is the set temperature threshold,
[0029] Furthermore, the lodging degree dfmj i Including: When i = 0, dfmj i The value of is equal to the lodging area at the current time; when i>0, dfmj i The calculation formula is:
[0030] dfj i =exp(χ1*dfmj i-1 +fs i *(χ2+χ3*jyl i +χ4*trhs i ));
[0031] Among them, fs i 、jyl i and trhs i Represented as the wind speed, rainfall and soil moisture content of the i-th day after the current time; dfmj i-1 It is expressed as the lodging area on the i-1th day after the current time; χ1, χ2, χ3 and χ4 represent the second lodging degree coefficient, the second wind speed coefficient, the third rainfall coefficient and the second soil moisture coefficient, respectively.
[0032] The harvesting suitability is sorted from high to low, the sorted harvesting suitability and the corresponding harvesting time are displayed, and the suitable harvesting time is displayed.
[0033] Furthermore, the process of obtaining the suitable harvesting time includes:
[0034] When the harvesting suitability is greater than the set minimum harvesting suitability, the harvesting time corresponding to the harvesting suitability is used as the suitable harvesting time; when the harvesting suitability of each day in the harvesting time interval is less than the set minimum harvesting suitability, the suitable harvesting time output is empty.
[0035] An information processing system based on the Internet of Things, comprising:
[0036] The first acquisition unit is used to acquire farmland images, and acquire the types of crops in the farmland and the current growth status of the crops according to the farmland images;
[0037] The second acquisition unit is used to acquire the harvesting time interval corresponding to the crop type in the farmland when the current growth state of the crop is mature.dq ,time dq +T zl ] in the harvesting environment data; among them, time dq Indicates the current time; T zl It is represented by the maximum harvesting time interval corresponding to the crop type in the farmland, T zl is a positive integer;
[0038] Harvesting suitability calculation unit: used to obtain harvesting suitability according to the harvesting environment data, and the calculation formula is:
[0039]
[0040] Among them, syd i cs It is expressed as the harvest suitability of the i-th day after the current time, and the value of i ranges from 0 to T zl ; When the value of i is 0, syd i cs Represents the harvest suitability at the current time; trhs i 、jyl i 、pjwd i 、fs i and dfmj i They respectively represent the soil moisture content, rainfall, average temperature, maximum wind speed and lodging area of the ith day after the current time; bzwd represents the standard maturity temperature value corresponding to the set crop type in the farmland; α1, α2, α3, α4 and α5 represent the first soil moisture coefficient, the first rainfall coefficient, the first temperature coefficient, the first wind speed coefficient and the first lodging degree coefficient; exp() represents an exponential function with a natural constant as the base;
[0041] Display unit: used to sort the harvesting suitability from high to low, display the sorted harvesting suitability and corresponding harvesting time, and display the suitable harvesting time.
[0042] Furthermore, the harvesting environment data includes: the current farmland soil moisture content, the current crop lodging area, the daily rainfall within the harvesting time interval, the average temperature, the light time, the light intensity and the maximum wind speed.
[0043] Compared with the prior art, the present invention has the following beneficial effects:
[0044] 1. The present invention constructs a crop growth status acquisition model, which can acquire multiple crop images for farmland images, and acquire the crop color features, crop leaf texture features, crop leaf shape features and crop leaf size features of the crop images, and identify the types of crops and the growth status of crops based on these features. This model acquires the types of farmland crops and the current growth status corresponding to the farmland images based on the corresponding crop types and growth status of different crop images. This model can avoid the impact of a single crop image abnormality on the detection of the types of farmland crops and the current growth status of the entire farmland, thereby improving the accuracy of the detection of the types of farmland crops and the current growth status, and providing a basis for the subsequent calculation of the optimal harvesting time of crops.
[0045] 2. The lodging of crops in farmland is affected not only by wind speed, but also by rainfall and soil moisture. The present invention obtains the lodging degree of crops every day in the harvesting time interval based on the daily rainfall, wind speed, soil moisture content and the lodging degree of the previous day in the harvesting time interval. The lodging degree not only affects the difficulty of harvesting but also affects the yield of crops. This method combines various factors that affect the lodging degree of crops, and can accurately obtain the predicted value of the lodging degree of crops, providing a data basis for the subsequent harvesting suitability.
[0046] 3. Weather factors not only affect the yield of farmland crops themselves, but also affect the harvesting efficiency of the machine. The present invention obtains the harvesting suitability of each day within the harvesting time interval based on the soil moisture content, rainfall, average temperature, maximum wind speed and lodging degree of each day within the harvesting time interval. This method can combine various weather factors with the soil moisture content and lodging degree of the farmland itself, and then accurately obtain the harvesting suitability, and obtain the appropriate harvesting time according to the harvesting suitability, providing a reference for the subsequent crop harvesting. BRIEF DESCRIPTION OF THE DRAWINGS
[0047] Figure 1 A flow chart of an information processing method based on the Internet of Things provided by an embodiment of the present invention;
[0048] Figure 2 A structural diagram of an information processing system based on the Internet of Things provided by an embodiment of the present invention;
[0049] Figure 3 A schematic diagram of the structure of a crop growth status acquisition model provided by an embodiment of the present invention;
[0050] Figure 4 A schematic diagram of the structure of a harvesting suitability calculation unit provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0052] The present invention provides an information processing method based on the Internet of Things. The method is applied to an information processing system based on the Internet of Things. For specific method flow charts and system structure diagrams, refer to Figure 1 and Figure 2 .
[0053] Embodiment 1
[0054] Reference Figure 1 S10 in the figure is applied to a first acquisition unit of an information processing system based on the Internet of Things.
[0055] Furthermore, this embodiment collects farmland images of farmland 1, and obtains the types of crops in the farmland and the current growth status of the crops based on the farmland images. The specific steps include: taking photos of the farmland by a drone every day to obtain the farmland images taken by the drone; inputting the farmland images into the crop growth status acquisition model to obtain the types of crops in the farmland and the current growth status; referring to Figure 3 The crop growth status acquisition model includes an input layer, a crop detection layer, a crop feature extraction layer, a crop feature fusion layer, a crop type acquisition layer, a crop growth status acquisition layer and an output layer;
[0056] The input layer is used to input the farmland image into the crop growth status acquisition model;
[0057] The crop detection layer is used to obtain N crop images according to the farmland image; N is an integer, and N can be 30, 40, 50, etc.; the crop feature extraction layer is used to obtain crop feature maps of N crop images according to the crop images; the crop feature extraction layer uses the crop ShuffleNetV2 submodel to extract the crop feature map of the crop image; the crop ShuffleNetV2 submodel is based on the ShuffleNetV2 model, and the output and input of each convolution layer are residually connected through a residual block, and a spatial attention mechanism is added after each residual block to obtain the final output of each residual block, and the final output is used as the input of the next convolution layer;
[0058] The crop feature fusion layer is used to perform feature fusion on the crop feature map corresponding to each crop image to obtain crop features corresponding to N crop images; the crop features include crop color features, crop leaf texture features, crop leaf shape features and crop leaf size features; the crop color features include crop leaf color features and crop fruit area color features; the crop type acquisition layer is used to obtain crop types corresponding to N crop images based on the crop features, and obtain the farmland crop types corresponding to the farmland image;
[0059] The crop growth status acquisition layer is used to acquire the current growth status of the crops according to the crop types in the farmland and the crop characteristics; the output layer is used to output the crop types in the farmland and the current growth status.
[0060] This step of the present embodiment constructs a crop growth status acquisition model, which can acquire multiple crop images for farmland images, and acquire crop feature maps of crop images through the crop ShuffleNetV2 sub-model, and then acquire crop color features, crop leaf texture features, crop leaf shape features and crop leaf size features, and identify the types of crops and the growth status of crops based on these features. The crop ShuffleNetV2 sub-model can help the crop ShuffleNetV2 sub-model dynamically select and adjust the importance of crop features through residual connections and spatial attention mechanisms, thereby enhancing the crop ShuffleNetV2 sub-model's ability to perceive key features. The crop growth status acquisition model can accurately acquire various features of crop images through the crop ShuffleNetV2 sub-model, and accurately acquire the types of crops and current growth status in farmland based on different characteristics of crops.
[0061] Furthermore, the crop type acquisition layer further comprises: when the crop types corresponding to the N crop images are different, constructing a type set (ZL1, ZL2, ..., ZL M ), where M represents the number of crop types identified by the N crop images, and ZL M It is represented by the number of the Mth crop species identified in the N crop images;
[0062] Get the maximum value ZL in the category set max ; When ZL max When / N is greater than the set category ratio threshold, the ZL max The corresponding crop type is used as the crop type of the farmland; the ZL maxThe crop features of the corresponding crop images are input into the crop growth status acquisition layer;
[0063] When ZL max / N is not greater than a set category ratio threshold, the crop detection layer reacquires N crop images; the category ratio threshold is obtained through historical data training.
[0064] Furthermore, when the number of times the crop detection layer reacquires N crop images reaches a set repetition threshold, the repetition threshold may be 3, 4, 5, etc.; the crop detection layer directly jumps to the output layer to output an abnormal crop type warning.
[0065] Occasionally, there are crops of different types in the farmland, and different types of crops have different corresponding maturity characteristics and maturity times. This step of the present embodiment can prevent the influence of abnormal types of crops on the subsequent identification of the current maturity state of the farmland during the planting of the farmland.
[0066] Furthermore, the crop growth status acquisition layer also includes: comparing each of the crop features received by the crop growth status acquisition layer with the crop maturity features corresponding to the crop types in the farmland to obtain feature similarity; when the feature similarity is greater than a set similarity threshold, it is indicated that the growth state corresponding to the crop feature is mature; the similarity threshold is obtained through training with a large amount of historical data; the number of the crop features with a mature growth state is obtained, recorded as sl cs ; When sl cs / sl js When the ratio of the crop to the mature state is greater than the set maturity state ratio threshold, the current growth state of the crop is mature. js It is represented by the number of the crop features received by the crop growth state acquisition layer; otherwise, the current growth state of the crop is immature; the maturity state ratio threshold is obtained through training with a large amount of historical data.
[0067] In this step of the present embodiment, each of the crop characteristics is compared with the crop maturity characteristics corresponding to the crop type in the farmland, and the growth state corresponding to each of the crop characteristics is obtained, thereby obtaining the current growth state of the crops in the farmland. This method can accurately obtain the current growth state of the crops, providing a basis for obtaining the suitability for harvesting later.
[0068] In order to verify the effectiveness of the crop growth status acquisition model in this embodiment, this embodiment collected 300 farmland images of different crops in different growth states, and input the 300 images into four models for detection. Each model was tested repeatedly 5 times, and the detection accuracy was statistically analyzed. The statistical results are shown in Table 1.
[0069] Table 1 Growth status detection comparison table
[0070] Repetitions Model 1 Model 2 Model 3 Model 4 1 95.33% 92.67% 92.33% 89.33% 2 94.67% 92.33% 91.67% 89.00% 3 95.00% 93.00% 91.33% 89.67% 4 94.33% 92.00% 92.00% 90.00% 5 94.67% 92.67% 91.67% 89.33%
[0071] Referring to Table 1, Model 1 in Table 1 is a crop growth status acquisition model proposed in the embodiment; Model 2 is to delete the crop type acquisition layer based on Model 1; Model 3 is to replace the crop ShuffleNetV2 sub-model with the ShuffleNetV2 model based on Model 1 to extract the crop feature map; Model 4 is to use the CNN model to detect the crop growth status. It can be seen from Table 1 that the crop growth status acquisition model proposed in this embodiment is higher than the other three models in 5 repeated tests. Therefore, the crop growth status acquisition model proposed in this embodiment can effectively improve the accuracy of the current growth status detection of the farmland.
[0072] Reference Figure 1 S20 in the figure is applied to a second acquisition unit of an information processing system based on the Internet of Things.
[0073] Furthermore, when the current growth state of the crop is immature, the drone continues to acquire images of the farmland; when the current growth state of the crop is mature, the drone acquires a harvesting time interval corresponding to the crop type in the farmland. dq ,time dq +T zl ] in the harvesting environment data; among them, time dq Indicates the current time; T zl It is represented by the maximum harvesting time interval corresponding to the crop type in the farmland, T zl is a positive integer; T zl It is obtained based on expert experience.
[0074] Furthermore, the harvesting environment data is acquired through Internet of Things sensors; the Internet of Things sensors include soil moisture sensors, intelligent meteorological sensors, satellite sensors and camera sensors; the harvesting environment data include farmland soil moisture content at the current time, crop lodging area at the current time, daily rainfall within the harvesting time interval, average temperature, light time, light intensity and maximum wind speed.
[0075] Furthermore, the daily average temperature, illumination time, illumination intensity and maximum wind speed within the harvesting time interval are obtained through intelligent meteorological sensors; the daily rainfall within the harvesting time interval is obtained through satellite sensors; the soil moisture content of the farmland at the current time is obtained through a soil moisture sensor; and the crop lodging area at the current time is obtained through a camera sensor.
[0076] In this embodiment, this step obtains various harvesting environment data within the harvesting time interval, and the harvesting environment data includes not only the weather data within the harvesting time interval but also the farmland's own data at the current time. This method provides a data basis for the subsequent harvesting suitability calculation.
[0077] Reference Figure 1 S30 in the above description is applied to a harvesting suitability calculation unit of an information processing system based on the Internet of Things. Figure 4 The harvesting suitability calculation unit includes a soil moisture calculation module, a lodging degree calculation module and a harvesting suitability calculation module; the soil moisture calculation module is used to obtain the soil moisture content of the farmland every day within the harvesting time interval according to the rainfall, average temperature, lighting time, lighting intensity and soil moisture at the current time in the harvesting environment data; the lodging degree calculation module is used to obtain the lodging degree of the farmland every day within the harvesting time interval according to the wind speed, rainfall in the harvesting environment data and the soil moisture calculated by the soil moisture calculation module; the harvesting suitability calculation module is used to obtain the harvesting suitability of the farmland every day within the harvesting time interval according to the harvesting environment data, the soil moisture calculated by the soil moisture calculation module and the lodging degree calculated by the lodging degree calculation module.
[0078] Furthermore, the daily soil moisture content in the field during the harvest time interval is trhs i The calculation process includes: trhs i It is expressed as the soil moisture content on the i-th day after the current time; when i = 0, trhs i It is expressed as the soil moisture content at the current time; when i>0, trhs i The calculation formula is:
[0079]
[0080] Among them, jyl i 、pjwd i 、gzsj i and gzqd i They represent the rainfall, average temperature, sunshine time and sunshine intensity of the ith day after the current time; trhs i-1 It is expressed as the soil moisture content on the i-1th day after the current time; β1, β2, β3 and β4 represent the second rainfall coefficient, the second average temperature coefficient, the light time coefficient and the light intensity coefficient respectively; the default values of β1, β2, β3 and β4 are all 0.25, which can also be changed according to the specific implementation situation in actual application; is the set temperature threshold, It can be 1, 2, 3, etc.
[0081] In this embodiment, the step obtains the soil moisture content of the farmland on the day based on the rainfall, average temperature, sunshine time and sunshine intensity of the day and the soil moisture content of the previous day. This method can accurately obtain the soil moisture content of the farmland every day during the harvest time interval, thereby providing a data basis for the subsequent calculation of the lodging degree and harvesting suitability.
[0082] Furthermore, the degree of lodging in the field every day during the harvest time interval dfmj i The calculation process includes: dfmj i It is expressed as the lodging degree of the i-th day after the current time; when i = 0, dfmj i The value of is equal to the lodging area at the current time; when i>0, dfmj i The calculation formula is:
[0083] dfj i =exp(χ1*dfmj i-1 +fs i *(χ2+χ3*jyl i +χ4*trhs i ));
[0084] Among them, fs i 、jyl i and trhs i Represented as the wind speed, rainfall and soil moisture content of the i-th day after the current time; dfmj i-1 It is expressed as the lodging area on the i-1th day after the current time; χ1, χ2, χ3 and χ4 represent the second lodging degree coefficient, the second wind speed coefficient, the third rainfall coefficient and the third soil moisture coefficient respectively; the default values of χ1, χ2, χ3 and χ4 are all 0.25, which can also be changed according to the specific implementation situation in actual application.
[0085] The lodging of crops in farmland is affected not only by wind speed, but also by rainfall and soil moisture. This embodiment obtains the lodging degree of crops every day in the harvesting time interval based on the daily rainfall, wind speed, soil moisture content and the lodging degree of the previous day in the harvesting time interval. The lodging degree not only affects the difficulty of harvesting but also affects the yield of crops. This method combines various factors that affect the lodging degree of crops, and can accurately obtain the predicted value of the lodging degree of crops, providing a data basis for the subsequent harvesting suitability.
[0086] Furthermore, the calculation formula for the harvesting suitability is:
[0087]
[0088] in, It is expressed as the harvest suitability of the i-th day after the current time, and the value of i ranges from 0 to T zl ; When the value of i is 0, Represents the harvest suitability at the current time; trhs i 、jyl i 、pjwd i 、fs i and dfmj i They respectively represent the soil moisture content, rainfall, average temperature, maximum wind speed and lodging degree of the i-th day after the current time; bzwd represents the standard maturity temperature value corresponding to the set type of crop in the farmland; α1, α2, α3, α4 and α5 represent the first soil moisture coefficient, the first rainfall coefficient, the first temperature coefficient, the first wind speed coefficient and the first lodging degree coefficient respectively; the values of α1, α2, α3, α4 and α5 are all 0.2 by default, and can be changed according to the specific implementation situation in actual application; exp() represents an exponential function with natural constant as the base.
[0089] Weather factors not only affect the yield of farmland crops themselves, but also affect the harvesting efficiency of the machine. In this embodiment, this step obtains the harvesting suitability for each day in the harvesting time interval based on the soil moisture content, rainfall, average temperature, maximum wind speed and lodging degree of each day in the harvesting time interval. This method can combine various weather factors with the soil moisture content and lodging degree of the farmland itself, and then accurately obtain the harvesting suitability.
[0090] Reference Figure 1 The S40 in the embodiment is applied to a display unit of an information processing system based on the Internet of Things.
[0091] Furthermore, the process of obtaining the suitable harvesting time includes:
[0092] When the harvesting suitability is greater than the set minimum harvesting suitability, the harvesting time corresponding to the harvesting suitability is used as the suitable harvesting time; different types of crops have different corresponding minimum harvesting suitabilities; the minimum harvesting suitability is obtained through training with a large amount of historical data; when the harvesting suitability of each day in the harvesting time interval is less than the set minimum harvesting suitability, the suitable harvesting time output is empty.
[0093] Furthermore, in this embodiment, it is identified that the current growth state of the crops in Farmland 1 is mature, and the corresponding maximum harvesting time interval is obtained as 7 days based on the type of crops in Farmland 1; in this embodiment, based on the harvesting environment data within the harvesting time interval, the harvesting suitability within the harvesting time interval is obtained, and the harvesting suitability is sorted from high to low, and the suitable harvesting time, the sorted harvesting suitability and the corresponding harvesting time are displayed. For specific results, refer to Table 2.
[0094] Table 2 Example 1 Harvest suitability and suitable harvesting schedule
[0095] Serial number Harvest time Harvesting suitability Is it the right time to harvest? 1 <![CDATA[time dq +2]]> 2.13 yes 2 <![CDATA[time dq +3]]> 2.05 yes 3 <![CDATA[time dq +4]]> 1.98 yes 4 <![CDATA[time dq ]]> 1.75 no 5 <![CDATA[time dq +1]]> 1.71 no 6 <![CDATA[time dq +7]]> 1.68 no 7 <![CDATA[time dq +6]]> 1.64 no 8 <![CDATA[time dq +5]]> 1.59 no
[0096] In this step of the present embodiment, the harvesting suitability of each day within the harvesting time interval is compared with the set minimum harvesting suitability, and the suitable harvesting time is obtained. This method can judge the suitable harvesting time by combining various harvesting environmental factors, and provide a reference for agricultural producers.
[0097] Embodiment 2
[0098] Reference Figure 1 S10 in the figure is applied to a first acquisition unit of an information processing system based on the Internet of Things.
[0099] Furthermore, the present embodiment collects farmland images of farmland 2, and obtains the types of crops in the farmland and the current growth status of the crops based on the farmland images. The specific steps include: obtaining farmland images taken by a drone; inputting the farmland images into a crop growth status acquisition model to obtain the types of crops and the current growth status; the crop growth status acquisition model includes an input layer, a crop detection layer, a crop feature extraction layer, a crop feature fusion layer, a crop type acquisition layer, a crop growth status acquisition layer and an output layer;
[0100] The input layer is used to input the farmland image into the crop growth status acquisition model;
[0101] The crop detection layer is used to obtain N crop images according to the farmland image;
[0102] The crop feature extraction layer is used to obtain crop feature maps of N crop images based on the crop images;
[0103] The crop feature extraction layer uses the crop ShuffleNetV2 sub-model to extract the crop feature map of the crop image;
[0104] The crop ShuffleNetV2 sub-model is based on the ShuffleNetV2 model, and the output and input of each convolutional layer are residually connected through a residual block, and a spatial attention mechanism is added after each residual block to obtain the final output of each residual block, and the final output is used as the input of the next convolutional layer;
[0105] The crop feature fusion layer is used to perform feature fusion on the crop feature map corresponding to each crop image to obtain crop features corresponding to N crop images; the crop features include crop color features, crop leaf texture features, crop leaf shape features and crop leaf size features;
[0106] The crop type acquisition layer is used to acquire the crop types corresponding to the N crop images according to the crop characteristics, and acquire the farmland crop types corresponding to the farmland image;
[0107] The crop growth status acquisition layer is used to acquire the current growth status of the crops according to the crop types in the farmland and the crop characteristics; the output layer is used to output the crop types in the farmland and the current growth status.
[0108] Furthermore, the crop type acquisition layer further comprises: when the crop types corresponding to the N crop images are different, constructing a type set (ZL1, ZL2, ..., ZL M ), where M represents the number of crop types identified by the N crop images, and ZL M It is represented by the number of the Mth crop species identified in the N crop images;
[0109] Get the maximum value ZL in the category set max ; When ZL max When / N is greater than the set category ratio threshold, the ZL max The corresponding crop type is used as the crop type of the farmland; the ZL max The crop features of the corresponding crop images are input into the crop growth status acquisition layer;
[0110] When ZL max / N is not greater than a set category ratio threshold, the crop detection layer reacquires N crop images; the category ratio threshold is obtained through historical data training.
[0111] Furthermore, when the number of times the crop detection layer reacquires N crop images reaches a set repetition threshold, the crop detection layer directly jumps to the output layer to output an abnormal crop type warning.
[0112] Furthermore, the crop growth status acquisition layer further includes: comparing each of the crop features received by the crop growth status acquisition layer with the crop maturity features corresponding to the crop types in the farmland to obtain feature similarity; when the feature similarity is greater than a set similarity threshold, it is indicated that the growth state corresponding to the crop feature is mature; obtaining the number of the crop features with a mature growth state, recorded as sl cs ;
[0113] When sl cs / sl js When the ratio of the crop to the mature state is greater than the set maturity state ratio threshold, the current growth state of the crop is mature. js It represents the number of the crop features received by the crop growth state acquisition layer; otherwise, the current growth state of the crop is immature.
[0114] Reference Figure 1 S20 in the figure is applied to a second acquisition unit of an information processing system based on the Internet of Things.
[0115] Furthermore, when the current growth state of the crop is mature, the harvesting time interval corresponding to the crop type in the farmland is obtained. dq ,time dq +T zl ] in the harvesting environment data; among them, time dq Indicates the current time; T zl It is represented by the maximum harvesting time interval corresponding to the crop type in the farmland, T zl is a positive integer; T zl It is obtained based on expert experience.
[0116] Furthermore, the harvesting environment data includes the current farmland soil moisture content, the current crop lodging area, the daily rainfall within the harvesting time interval, the average temperature, the light time, the light intensity and the maximum wind speed.
[0117] Furthermore, the daily rainfall, average temperature, sunshine time, sunshine intensity and maximum wind speed within the harvesting time interval are obtained through weather forecast.
[0118] Reference Figure 1 S30 in the above description is applied to a harvesting suitability calculation unit of an information processing system based on the Internet of Things. Figure 4The harvesting suitability calculation unit includes a soil moisture calculation module, a lodging degree calculation module and a harvesting suitability calculation module; the soil moisture calculation module is used to obtain the soil moisture content of the farmland every day within the harvesting time interval according to the rainfall, average temperature, lighting time, lighting intensity and soil moisture at the current time in the harvesting environment data; the lodging degree calculation module is used to obtain the lodging degree of the farmland every day within the harvesting time interval according to the wind speed, rainfall in the harvesting environment data and the soil moisture calculated by the soil moisture calculation module; the harvesting suitability calculation module is used to obtain the harvesting suitability of the farmland every day within the harvesting time interval according to the harvesting environment data, the soil moisture calculated by the soil moisture calculation module and the lodging degree calculated by the lodging degree calculation module.
[0119] Furthermore, the daily soil moisture content in the field during the harvest time interval is trhs i The calculation process includes: trhs i It is expressed as the soil moisture content on the i-th day after the current time; when i = 0, trhs i It is expressed as the soil moisture content at the current time; when i>0, trhs i The calculation formula is:
[0120]
[0121] Among them, jyl i 、pjwd i 、gzsj i and gzqd i They represent the rainfall, average temperature, sunshine time and sunshine intensity of the ith day after the current time; trhs i-1 It is expressed as the soil moisture content on the i-1th day after the current time; β1, β2, β3 and β4 represent the second rainfall coefficient, the second average temperature coefficient, the light time coefficient and the light intensity coefficient respectively; is the set temperature threshold,
[0122] Furthermore, the degree of lodging in the field every day during the harvest time interval dfmj i The calculation process includes: dfmj i It is expressed as the lodging degree of the i-th day after the current time; when i = 0, dfmj i The value of is equal to the lodging area at the current time; when i>0, dfmj i The calculation formula is:
[0123] dfj i =exp(χ1*dfmj i-1 +fs i*(χ2+χ3*jyl i +χ4*trhs i ));
[0124] Among them, fs i 、jyl i and trhs i Represented as the wind speed, rainfall and soil moisture content of the i-th day after the current time; dfmj i-1 It is expressed as the lodging area on the i-1th day after the current time; χ1, χ2, χ3 and χ4 represent the second lodging degree coefficient, the second wind speed coefficient, the third rainfall coefficient and the third soil moisture coefficient respectively; the default values of χ1, χ2, χ3 and χ4 are all 0.25, which can also be changed according to the specific implementation situation in actual application.
[0125] Furthermore, the calculation formula for the harvesting suitability is:
[0126]
[0127] in, It is expressed as the harvest suitability of the i-th day after the current time, and the value of i ranges from 0 to T zl ; When the value of i is 0, Represents the harvest suitability at the current time; trhs i 、jyl i 、pjwd i 、fs i and dfmj i They respectively represent the soil moisture content, rainfall, average temperature, maximum wind speed and lodging degree of the i-th day after the current time; bzwd represents the standard maturity temperature value corresponding to the set type of crop in the farmland; α1, α2, α3, α4 and α5 represent the first soil moisture coefficient, the first rainfall coefficient, the first temperature coefficient, the first wind speed coefficient and the first lodging degree coefficient respectively; the values of α1, α2, α3, α4 and α5 are all 0.2 by default, and can be changed according to the specific implementation situation in actual application; exp() represents an exponential function with natural constant as the base.
[0128] Reference Figure 1 The S40 in the embodiment is applied to a display unit of an information processing system based on the Internet of Things.
[0129] Furthermore, the process of obtaining the suitable harvesting time includes:
[0130] When the harvesting suitability is greater than the set minimum harvesting suitability, the harvesting time corresponding to the harvesting suitability is used as the suitable harvesting time; different types of crops have different corresponding minimum harvesting suitabilities;
[0131] The minimum harvesting suitability is obtained through training with a large amount of historical data;
[0132] When the harvesting suitability of each day in the harvesting time interval is less than the set minimum harvesting suitability, the suitable harvesting time output is empty.
[0133] Furthermore, in this embodiment, it is identified that the current growth state of the crops in Farmland 2 is mature, and the corresponding maximum harvesting time interval is obtained as 13 days based on the types of crops in Farmland 2; this embodiment obtains the harvesting suitability within the harvesting time interval based on the harvesting environment data within the harvesting time interval, sorts the harvesting suitability from high to low, and displays the suitable harvesting time, the sorted harvesting suitability and the corresponding harvesting time. For specific results, refer to Table 3.
[0134] Table 3 Example 2 Harvest suitability and suitable harvesting schedule
[0135]
[0136] This embodiment constructs a crop growth state acquisition model, which can acquire multiple crop images for farmland images, and acquire crop color features, crop leaf texture features, crop leaf shape features and crop leaf size features of the crop images, and identify the types of crops and the growth status of crops based on these features. This model acquires the types of farmland crops and the current growth status corresponding to the farmland images based on the types and growth status of the crops corresponding to different crop images. This model can avoid the impact of a single crop image abnormality on the detection of the types of farmland crops and the current growth status of the entire farmland, thereby improving the accuracy of the detection of the types of farmland crops and the current growth status, and thus providing a basis for the subsequent calculation of the optimal harvesting time of crops. In addition to being affected by wind speed, the lodging of crops in farmland is also affected by rainfall and soil moisture. This embodiment acquires the lodging degree of crops every day in the harvesting time interval based on the daily rainfall, wind speed, soil moisture content and the lodging degree of the previous day in the harvesting time interval. The degree of lodging not only affects the difficulty of harvesting but also affects the yield of crops. This method combines various factors that affect the degree of lodging of crops, and can accurately obtain the predicted value of the degree of lodging of crops, providing a data basis for the subsequent suitability of harvesting. Weather factors not only affect the yield of farmland crops themselves but also affect the harvesting efficiency of the machine. This embodiment obtains the suitability of harvesting every day within the harvesting time interval based on the soil moisture content, rainfall, average temperature, maximum wind speed and degree of lodging every day within the harvesting time interval. This method can combine various weather factors with the soil moisture content and degree of lodging of the farmland itself, and then accurately obtain the suitability of harvesting, and obtain the suitable harvesting time according to the harvesting suitability, providing a reference for the subsequent harvesting of crops.
[0137] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. An information processing method based on the Internet of Things, characterized in that: include: Acquire farmland images, and acquire the types of crops in the farmland and the current growth status of the crops based on the farmland images; When the current growth state of the crop is mature, obtain the harvest time interval corresponding to the crop type in the farmland [time dq ,time dq +T zl ] in the harvesting environment data; among them, time dq Indicates the current time; T zl It is represented by the maximum harvesting time interval corresponding to the crop type in the farmland, T zl is a positive integer; According to the harvesting environment data, the harvesting suitability is obtained. The specific process includes: according to the harvesting environment, the harvesting suitability of each day within the harvesting time interval is obtained. The calculation formula of the harvesting suitability is: in, It is expressed as the harvest suitability of the i-th day after the current time, and the value of i ranges from 0 to T zl ; When the value of i is 0, Represents the harvest suitability at the current time; trhs i 、jyl i 、pjwd i 、fs i and dfmj i They respectively represent the soil moisture content, rainfall, average temperature, maximum wind speed and lodging degree of the ith day after the current time; bzwd represents the standard maturity temperature value corresponding to the set crop type in the farmland; α1, α2, α3, α4 and α5 represent the first soil moisture coefficient, the first rainfall coefficient, the first temperature coefficient, the first wind speed coefficient and the first lodging degree coefficient; exp() represents an exponential function with a natural constant as the base; The harvesting suitability is sorted from high to low, the sorted harvesting suitability and the corresponding harvesting time are displayed, and the suitable harvesting time is displayed.
2. According to the information processing method based on the Internet of Things according to claim 1, it is characterized in that: The acquiring of farmland images and acquiring the types of crops and the current growth status of the crops according to the farmland images include: Acquire a farmland image taken by a drone; input the farmland image into a crop growth status acquisition model to acquire the type of crops in the farmland and the current growth status; the crop growth status acquisition model includes an input layer, a crop detection layer, a crop feature extraction layer, a crop feature fusion layer, a crop type acquisition layer, a crop growth status acquisition layer and an output layer; the input layer is used to input the farmland image into the crop growth status acquisition model; The crop detection layer is used to obtain N crop images according to the farmland image; The crop feature extraction layer is used to obtain crop feature maps of N crop images based on the crop images; the crop feature fusion layer is used to perform feature fusion on the crop feature maps corresponding to each crop image to obtain crop features corresponding to the N crop images; the crop features include crop color features, crop leaf texture features, crop leaf shape features and crop leaf size features; The crop type acquisition layer is used to acquire the crop types corresponding to the N crop images according to the crop characteristics, and acquire the farmland crop types corresponding to the farmland image; The crop growth status acquisition layer is used to acquire the current growth status of the crops according to the crop types in the farmland and the crop characteristics; the output layer is used to output the crop types in the farmland and the current growth status.
3. The information processing method based on the Internet of Things according to claim 2 is characterized in that: The crop type acquisition layer further includes: when the crop types corresponding to the N crop images are different, constructing a type set (ZL1, ZL2, ..., ZL M ), where M represents the number of crop types identified by the N crop images, and ZL M It is represented by the number of the Mth crop species identified in the N crop images; Get the maximum value ZL in the category set max ; When ZL max When / N is greater than the set category ratio threshold, the ZL max The corresponding crop type is used as the crop type of the farmland; the ZL max The crop features of the corresponding crop images are input into the crop growth status acquisition layer; When ZL max When / N is not greater than the set category proportion threshold, the crop detection layer reacquires N crop images.
4. The information processing method based on the Internet of Things according to claim 2 is characterized in that: The crop growth state acquisition layer further comprises: comparing each of the crop features received by the crop growth state acquisition layer with the crop maturity features corresponding to the crop types in the farmland to obtain feature similarity; When the feature similarity is greater than the set similarity threshold, it is indicated that the growth state corresponding to the crop feature is mature; the number of the crop features whose growth state is mature is obtained, which is recorded as sl cs ; When sl cs / sl js When the ratio of the crop to the mature state is greater than the set maturity state ratio threshold, the current growth state of the crop is mature. js It represents the number of the crop features received by the crop growth state acquisition layer; otherwise, the current growth state of the crop is immature.
5. The information processing method based on the Internet of Things according to claim 1 is characterized in that: The harvesting environment data include: the current farmland soil moisture content, the current crop lodging area, the daily rainfall within the harvesting time interval, the average temperature, the light time, the light intensity and the maximum wind speed.
6. The information processing method based on the Internet of Things according to claim 1 is characterized in that: The soil moisture content trhs i Including: When i = 0, trhs i It is expressed as the soil moisture content at the current time; when i>0, trhs i The calculation formula is: Among them, jyl i 、pjwd i 、gzsj i and gzqd i They represent the rainfall, average temperature, sunshine time and sunshine intensity of the ith day after the current time; trhs i-1 It is represented as the soil moisture content on the i-1th day after the current time; β1, β2, β3 and β4 represent the second soil moisture coefficient, the second rainfall coefficient, the second average temperature coefficient and the light coefficient respectively; is the set temperature threshold, 7. The information processing method based on the Internet of Things according to claim 1 is characterized in that: The lodging degree dfmj i Including: When i = 0, dfmj i The value of is equal to the lodging area at the current time; when i>0, dfmj i The calculation formula is: dfmj i =exp(χ1*dfmj i-1 +fs i *(x2+x3*jyl i +x4*trhs i )); Among them, fs i 、jyl i and trhs i Represented as the wind speed, rainfall and soil moisture content of the i-th day after the current time; dfmj i-1 It is expressed as the lodging area on the i-1th day after the current time; χ1, χ2, χ3 and χ4 represent the second lodging degree coefficient, the second wind speed coefficient, the third rainfall coefficient and the third soil moisture coefficient, respectively.
8. The information processing method based on the Internet of Things according to claim 1 is characterized in that: The process of obtaining the suitable harvesting time includes: When the harvesting suitability is greater than the set minimum harvesting suitability, the harvesting time corresponding to the harvesting suitability is used as the suitable harvesting time; when the harvesting suitability of each day in the harvesting time interval is less than the set minimum harvesting suitability, the suitable harvesting time output is empty.
9. An information processing system based on the Internet of Things, characterized in that: include: The first acquisition unit is used to acquire farmland images, and acquire the types of crops in the farmland and the current growth status of the crops according to the farmland images; The second acquisition unit is used to acquire the harvesting time interval corresponding to the crop type in the farmland when the current growth state of the crop is mature. dq ,time dq +T zl ] in the harvesting environment data; among them, time dq Indicates the current time; T zl It is represented by the maximum harvesting time interval corresponding to the crop type in the farmland, T zl is a positive integer; Harvesting suitability calculation unit: used to obtain harvesting suitability according to the harvesting environment data, and the calculation formula is: in, It is expressed as the harvest suitability of the i-th day after the current time, and the value of i ranges from 0 to T zl ; When the value of i is 0, Represents the harvest suitability at the current time; trhs i 、jyl i 、pjwd i 、fs i and dfmj i They respectively represent the soil moisture content, rainfall, average temperature, maximum wind speed and lodging degree of the ith day after the current time; bzwd represents the standard maturity temperature value corresponding to the set crop type in the farmland; α1, α2, α3, α4 and α5 represent the first soil moisture coefficient, the first rainfall coefficient, the first temperature coefficient, the first wind speed coefficient and the first lodging degree coefficient; exp() represents an exponential function with a natural constant as the base; Display unit: used to sort the harvesting suitability from high to low, display the sorted harvesting suitability and corresponding harvesting time, and display the suitable harvesting time.
10. The information processing system based on the Internet of Things according to claim 9, characterized in that: The harvesting environment data include: the current farmland soil moisture content, the current crop lodging area, the daily rainfall within the harvesting time interval, the average temperature, the light time, the light intensity and the maximum wind speed.
Citation Information
Patent Citations
Meteorological disaster early warning system and method based on agricultural Internet of Things
CN116011812A
Remote sensing recognition system of crop planting structure
CN117789023A
Equipment integrated control system and method based on smart agriculture
CN118503668A
Agricultural machinery harvesting route planning and processing device based on crop mature period prediction
CN218381030U
Polyurethane roof waterproofing material manufacturing method with excellent eco-friendliness
KR1020250001236A
Cited By
Combined control system of unmanned aerial vehicle and ground agricultural machinery
CN120993943A
