CN114742288A - Intelligent prediction method for chenopodium quinoa ear-sprouting yield - Google Patents

Abstract

The invention provides an intelligent prediction method for the ear-sprouting yield of quinoa, which belongs to the technical field of yield prediction and comprises the following steps: collecting environmental information such as temperature, humidity, illumination time and illumination intensity in the planting area, detecting nutrient information of soil in the planting area, and combining the nutrient information and the environmental information into a supply system. Scanning quinoa particles in the planting area, acquiring the size parameter of the current quinoa particles, deducing the number of the quinoa particles in the planting area, and calculating the total quantity of the quinoa in the planting area according to the number and the size parameter. And generating a checking coefficient according to the empirical coefficient of the chenopodium quinoa yield and the supply system, and predicting the target yield of the planting area according to the checking coefficient and the chenopodium quinoa total amount. The target yield obtained by the intelligent prediction method for the chenopodium quinoa ear-sprouting yield is based on actual starting, the data result is accurate, the reference is strong, and powerful data support is provided for the development of agriculture.

Description

An Intelligent Prediction Method for Quinoa Heading Yield

technical field

The invention belongs to the technical field of yield prediction, and more particularly relates to an intelligent prediction method for quinoa heading yield.

Background technique

The cultivation area of quinoa in the three northeastern provinces is expanding year by year. In this area, planting is generally carried out in early May and harvested in mid or late September. Traditional quinoa is generally planted in a single ridge and row, with a ridge width ranging from 60cm-65cm. The quinoa plant is tall, with well-developed lateral branches and high nutritional value.

The output of agricultural products is closely related to people's daily life. Using the past output data and historical data of a series of influencing factors, the trend of the data itself and the influence relationship and degree of each factor are fully excavated, and certain methods and skills are used to predict future output. Forecasting and forecasting will help countries, producers and consumers to better judge the economic situation and make correct decisions.

Since there are many quinoa ears growing on the quinoa seedlings, and each quinoa ear has multiple approximately spherical quinoa particles, which causes a lot of troubles in the yield prediction of the quinoa planting area. The existing yield prediction methods Most of the judgments are made based on the variety and environmental factors. Since there is no combination of the planted quinoa grains, the final data is less referable and the data reliability is not high.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide an intelligent prediction method for quinoa heading yield, which aims to solve the problems of poor data reference and low data reliability when quinoa yield is predicted.

In order to achieve the above object, the technical scheme adopted in the present invention is: provide an intelligent prediction method for quinoa heading yield, including:

Collect environmental information such as temperature, humidity, light time and light intensity in the planting area, detect the nutrient information of the soil in the planting area, and combine the nutrient information and the environmental information into a supply system;

Scan the quinoa grains in the planting area, obtain the size parameter of the quinoa grains at present, infer the number of the quinoa grains in the planting area, and calculate from the quantity and the size parameter The total amount of quinoa in the planting area;

A calibration coefficient is generated according to the empirical model of quinoa yield and the supply system, and the target yield of the planting area is predicted by the calibration coefficient and the total amount of quinoa.

In a possible implementation manner, the scanning of the quinoa particles in the planting area, and obtaining the current size parameters of the quinoa particles include:

The planting area is divided into a plurality of blocks, and the quinoa particles in the plurality of blocks are sequentially scanned by the collection device;

The acquisition device transmits the scanned information to the upper computer, and the upper computer determines the size parameter according to the received information.

In a possible implementation manner, the step of sequentially scanning the quinoa particles in the block by the collection device includes:

The collection device is stabilized on one side of the corresponding block at a specific angle;

The acquisition device scans the quinoa particles through pictures, videos and structured light.

In a possible implementation, the obtaining the current size parameters of the quinoa particles includes:

The collection device obtains a plurality of characteristic points on the edge of the quinoa grain, and determines the size parameters such as the volume of the quinoa grain through the plurality of the characteristic points.

In a possible implementation, the inferring the number of the quinoa grains in the planting area includes:

The host computer determines the number of quinoa ears on the quinoa seedlings and the size of the quinoa ears through methods such as pictures and videos, and the host computer infers the number of the quinoa particles on the quinoa ears. , the number of the quinoa particles obtained in each of the blocks is added to finally determine the number.

In a possible implementation, the host computer deduces that the number of the quinoa particles on the quinoa ear includes:

According to the variety of quinoa, determine the number of the quinoa particles on the side of the quinoa ear facing the collecting device, and infer the total amount of the quinoa on the quinoa ear according to the size of the quinoa ear Number of quinoa grains.

In a possible implementation manner, the generation of the calibration coefficient according to the empirical model of quinoa yield and the supply system includes:

Corresponding influence factors are set for each component in the supply system, the empirical model includes empirical parameters for each component, and the value of each component in the supply system is compared with the value of each component in the empirical parameter. Compare the values of , and obtain the calibration coefficient from the empirical model according to the influence factor.

In a possible implementation manner, the value of each component in the supply system is compared with the value of each component in the empirical parameter, and the calibration is obtained from the empirical model according to the influence factor. Kernel coefficients include:

Predict the supply system of each interval from the current time point of quinoa to the harvest period, compare the predicted supply system with the empirical parameters, and combine the compared results with the impact factors to determine the range of each interval. The interval parameters are determined, and the calibration coefficient is finally obtained from the determined interval parameters.

In a possible implementation manner, the comparison result is combined with the influence factor to determine the interval parameters in each of the interval segments, and the calibration coefficient is finally obtained from the determined plurality of the interval parameters include:

Differentiate the value of each component in the supply system with the value of the corresponding component in the empirical parameter within the same interval, and compare the results of multiple differences with the corresponding influencing factors. Multiply, add the multiplied results to obtain the coefficient deviation, add the coefficient deviation to the empirical model to obtain the interval parameter; multiply the plurality of interval parameters to obtain the calibration coefficient .

In a possible implementation manner, the prediction of the target yield of the planting area by using the calibration coefficient and the total amount of quinoa includes:

The target yield is obtained by multiplying the calibration coefficient by the total amount of quinoa.

The beneficial effect of the intelligent prediction method for quinoa heading yield provided by the present invention is: compared with the prior art, the intelligent prediction method for quinoa heading yield of the present invention firstly includes temperature, humidity, illumination time, illumination intensity, etc. in the collection and planting area. environmental information, and the nutrient information of the soil in the planting area needs to be detected. The quinoa grains in the planting area were scanned to infer the number of quinoa grains, and the total amount of quinoa in the planting area was calculated from the number and the size parameters of the quinoa grains. The calibration coefficient is generated according to the empirical model and the supply system, and the target yield is predicted according to the calibration coefficient and the total amount of quinoa.

In this application, environmental information such as temperature, humidity, light time, and light intensity, and soil nutrient information form a supply system. The supply system is a necessary condition for the growth of quinoa and other plants, and an important reference factor for predicting the yield of quinoa. By scanning and inferring quinoa grains, the size parameters and quantities of quinoa grains were determined, thereby providing accurate data support for yield prediction. The supply system needs to generate a calibration coefficient according to the previous empirical model. The target yield obtained through the calibration coefficient and the total amount of quinoa is based on the actual situation. The data results are more accurate and have strong reference, which provides strong data for the development of agriculture. support.

Description of drawings

In order to illustrate the technical solutions in the embodiments of the present invention more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only for the present invention. In some embodiments, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without any creative effort.

FIG. 1 is a flowchart of an intelligent prediction method for quinoa heading yield provided by an embodiment of the present invention.

Detailed ways

In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention.

Referring to Fig. 1, the intelligent prediction method for quinoa heading yield provided by the present invention will now be described. Intelligent prediction methods for quinoa heading yield, including:

Collect environmental information such as temperature, humidity, light time and light intensity in the planting area, detect the nutrient information of the soil in the planting area, and combine the nutrient information and environmental information into a supply system.

Scan the quinoa grains in the planting area, obtain the size parameters of the current quinoa grains, infer the number of quinoa grains in the planting area, and calculate the total amount of quinoa in the planting area from the quantity and size parameters.

According to the empirical model of quinoa yield and the supply system, the calibration coefficient is generated, and the target yield of the planting area is predicted by the calibration coefficient and the total amount of quinoa.

The beneficial effect of the intelligent prediction method for quinoa heading yield provided by the present invention is: compared with the prior art, the intelligent prediction method for quinoa heading yield of the present invention firstly includes temperature, humidity, illumination time, illumination intensity, etc. in the collection and planting area. environmental information, and the nutrient information of the soil in the planting area needs to be detected. The quinoa grains in the planting area were scanned to infer the number of quinoa grains, and the total amount of quinoa in the planting area was calculated from the number and the size parameters of the quinoa grains. The calibration coefficient is generated according to the empirical model and the supply system, and the target yield is predicted according to the calibration coefficient and the total amount of quinoa.

In this application, environmental information such as temperature, humidity, light time, and light intensity, and soil nutrient information form a supply system. The supply system is a necessary condition for the growth of quinoa and other plants, and an important reference factor for predicting the yield of quinoa. By scanning and inferring quinoa grains, the size parameters and quantities of quinoa grains were determined, thereby providing accurate data support for yield prediction. The supply system needs to generate a calibration coefficient according to the previous empirical model. The target yield obtained through the calibration coefficient and the total amount of quinoa is based on the actual situation. The data results are more accurate and have strong reference, which provides strong data for the development of agriculture. support.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, the quinoa particles in the planting area are scanned to obtain the size parameters of the current quinoa particles including:

The planting area is divided into multiple blocks, and the quinoa grains in the multiple blocks are scanned by the collection equipment in turn.

The acquisition device transmits the scanned information to the upper computer, and the upper computer determines the size parameters according to the received information.

Since there are many areas where quinoa is grown, and there are many quinoa grains grown on each quinoa seedling, this has caused many difficulties in the scanning of quinoa grains. Will sway with the wind, making scanning quinoa grains a near-impossible task. Since the computing power of the current chip has been significantly improved, in order to complete the work of determining the size in a timely and efficient manner, the planting area needs to be divided first. The current method is to divide the planting area into multiple blocks of equal area in a certain order. , multiple blocks are arranged consecutively in a certain order. It should be noted that in order to collect quinoa size information from a specific angle in a specific block, it is usually necessary to use equipment such as drones. The coverage area of the acquisition equipment carried by the drone is limited, so when actually dividing the blocks The area of each block needs to be determined according to the specific model and performance adaptability of the acquisition equipment.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, sequentially scanning the quinoa particles in the block by the collection device includes:

The acquisition device is stabilized on one side of the corresponding block at a specific angle.

The acquisition equipment scans the quinoa grains through pictures, videos and structured light.

The acquisition equipment is used to obtain the three-dimensional information of the quinoa particles, and the drone is used to stabilize the acquisition equipment in a specific position of the corresponding block. The information acquired by the acquisition equipment is transmitted to the upper computer through the wireless network, and the powerful data processing capability of the upper computer is used. After completing the determination of the size of quinoa particles, when the amount of data collected is large, the information collected by the collection equipment can be backed up first, and then processed in sequence.

The collection device can collect multiple high-resolution pictures, and the multiple high-resolution pictures are taken in a certain order, and the specific size parameters of the quinoa particles can be obtained by synthesizing multiple pictures at different times. At this time, it is necessary to ensure the stability of the distance between the acquisition device and the quinoa particles to avoid the problem of image enlargement and reduction. The collection device can also collect a piece of video information, and the host computer determines the size of the quinoa particles photographed therein through the video information.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, obtaining the size parameters of the current quinoa particles includes:

The acquisition device obtains multiple feature points on the edge of the quinoa particle, and determines the size parameters such as the volume of the quinoa particle through the multiple feature points.

It should be pointed out that there are many quinoa particles on the quinoa seedlings. If each quinoa particle is scanned with high precision and all directions, it will place very high requirements on the accuracy of the collection equipment and the processing capacity of the upper computer. And adopting the above method will result in lower processing efficiency. In order to solve the above problems, and quinoa has an approximate spherical structure, it is only necessary to determine several important characteristic points of quinoa particles to obtain the size parameters of quinoa. Therefore, in the actual application of the collection equipment, only a few points on the edge of the quinoa need to be determined, and the center and volume of the quinoa particles can be determined through these points. Therefore, it is necessary to check the obtained volume according to the corresponding variety, that is, to adjust it appropriately to be close to the real volume.

In practical applications, the acquisition device emits a detection wave to the block, and the detection wave can be 3D structured light. The acquisition device scans the quinoa particles and detects the feature points of several different positions of the quinoa particles, so as to calculate the quinoa The volume of wheat kernels.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, it is inferred that the number of quinoa grains in the planting area includes:

The host computer determines the number of quinoa ears on the quinoa seedlings and the size of the quinoa ears through methods such as pictures and videos. The number of particles is added to finally determine the number.

There are multiple quinoa ears growing on a quinoa seedling, and each quinoa ear has multiple quinoa grains. The whole quinoa ear can be approximately regarded as a conical structure. All quinoa grains are scanned, which will undoubtedly increase the accuracy of the final yield prediction. However, this method firstly requires higher requirements for the collection equipment. At the same time, due to the occlusion between the quinoa ears, some quinoa grains cannot be detected. , which makes full detection impossible.

In order to improve the efficiency of the whole operation, several quinoa grains at specific positions can be selected on each quinoa ear, and according to the size of the quinoa ear, the number of quinoa grains on a quinoa ear can be calculated. The method can calculate the total amount of quinoa particles on the quinoa seedlings without scanning all the quinoa particles, thus greatly reducing the workload of data processing.

In order to solve the above problems, based on the growth characteristics of quinoa seedlings, the collection device of the present application scans quinoa grains located at a specific position on a quinoa ear. The radius of the quinoa grains on the quinoa ear gradually increases from the bottom to the top, so the collection device in this application will scan the quinoa grains at the top, middle and bottom, and then calculate the size of the quinoa grains at other positions , and finally determine the yield of the quinoa ear.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, the number of quinoa grains on the quinoa ear inferred by the host computer includes:

According to the variety of quinoa, the number of quinoa grains on the side of the quinoa ear facing the collection device is determined, and the total number of quinoa grains on the quinoa ear is deduced according to the size of the quinoa ear.

There will be differences in the size of the quinoa ears at different positions on a quinoa seedling. In order to achieve a more accurate estimation of the quinoa particles of the entire quinoa seedling, after determining the approximate size of the quinoa particles, the host computer obtains the quinoa particles according to the collection equipment. Information According to the experience and the variety of quinoa planted, the number of quinoa seedlings can be roughly calculated, and the yield can be calculated by multiplying the calculated number of quinoa grains by the detected volume of quinoa grains.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, generating the calibration coefficient according to the empirical model of quinoa yield and the supply system includes:

Set the corresponding influence factors for each component in the supply system. The empirical model includes the empirical parameters for each component. The calibration coefficients are obtained from the empirical model.

For the convenience of explanation, it is assumed that only the factors affecting the growth of quinoa are temperature, humidity, light time, light intensity and nutrient information. The coefficients of temperature, humidity, light time, light intensity and nutrient information are initially set respectively. are 0.2, 0.1, 0.3, 0.2 and 0.2. According to past experience, when there is a difference between the actual planted quinoa and the value corresponding to the historical record, that is, the empirical model, take the temperature as an example. For example, when the actual temperature is 5° higher than the corresponding value in the empirical model , 5 times 0.2 and finally get 1, then when determining the actual calibration coefficient, it is necessary to consider adding 1 on the basis of the original empirical model. Note that this method is only used to briefly illustrate the calculation idea. Specifically, when the temperature is too high, the yield of quinoa will decrease.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, the value of each component in the supply system is compared with the value of each component in the empirical parameters, and the calibration coefficient is obtained from the empirical model according to the impact factor, including: :

Predict the supply system of each interval from the current time point of quinoa to the harvest period, compare the predicted supply system with the empirical parameters, and determine the interval parameters in each interval by combining the results of the comparison with the impact factors. The multiple interval parameters of , and finally get the calibration coefficient.

In this application, in order to accurately predict the target yield, it is necessary to predict from the current prediction date to the mature supply system of quinoa. At this time, it is necessary to cooperate with the meteorological department to investigate the temperature, humidity, light time and light intensity of each interval. Then, the predicted supply system is compared with the temperature, humidity, light intensity and light time corresponding to the empirical model, so as to provide data support for the determination of target yield.

There are differences in the rate of volume change of quinoa grains in different periods. In order to accurately predict the final quinoa yield, it is necessary to determine which stage the current quinoa grains are in, and to predict the supply system in the future interval. . First of all, we will know the growth rate of the same variety of quinoa in an approximate external environment by consulting materials and other means. This growth rate can be used as an empirical model. Then, based on experience and related experiments, we can determine the impact of each factor in the supply system on the target yield. coefficient.

In actual operation, it is found that the nutrient information is easy to control, but the temperature, humidity, light intensity and light time cannot be accurately grasped, especially the actual temperature. When the temperature changes greatly within a period of time after the detection is completed, it will Lead to inaccurate forecast of target yield.

In order to solve the above problems, taking temperature as an example, in practical application, not only the change of temperature compared with the reference standard during the interval from sowing to the completion of testing, but also the change of temperature from the completion of testing to the end of growth should be considered. Happening.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, the comparison result is combined with the influence factor to determine the interval parameters in each interval, and the calibration coefficient is finally obtained from the determined interval parameters, including :

Differentiate the value of each component in the supply system with the value of the corresponding component in the empirical parameters within the same interval, multiply the results of multiple differences with the corresponding influence factors, and multiply the multiplied results. The results are added to obtain the coefficient deviation, and the coefficient deviation is added to the empirical model to obtain the interval parameters; the calibration coefficients are obtained by multiplying multiple interval parameters.

In this application, the quinoa particles need to be tested multiple times, and the growth rate of the volume of the quinoa particles in different periods will be different. Therefore, in actual operation, a reference standard needs to be set first, and the reference standard is the same variety of quinoa. The previous yield changes of wheat and the supply system are similar. The reference standard contains the parameters of quinoa in different intervals. By comparing the actual supply system with the reference standard, that is, the empirical model, the calibration coefficient will finally be obtained. The kernel coefficient and total quinoa yields the fitted target yield. The growth rate of quinoa particles is divided into several stages, and the influence of the supply system on quinoa in each stage will be different. Therefore, in the actual analysis, the calibration parameters need to be comprehensively considered in several intervals.

Taking temperature as an example, since the change of temperature in each interval will affect the growth of quinoa, when determining the target yield, it is necessary to add the influences of each interval, and then obtain a coefficient deviation.

In some embodiments of the intelligent prediction method for quinoa heading yield provided by the present application, the target yield of the planting area is predicted by the calibration coefficient and the total amount of quinoa including:

Multiply the calibration factor by the total amount of quinoa to get the target yield.

After the collection equipment determines the current volume of quinoa, the final target yield can be obtained by multiplying the determined calibration coefficient by the total amount of quinoa.

The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention shall be included in the protection scope of the present invention. Inside.

Claims (10)

1. the intelligent prediction method of quinoa heading yield, is characterized in that, comprises: Collect environmental information such as temperature, humidity, light time and light intensity in the planting area, detect the nutrient information of the soil in the planting area, and combine the nutrient information and the environmental information into a supply system; Scan the quinoa grains in the planting area, obtain the size parameter of the quinoa grains at present, infer the number of the quinoa grains in the planting area, and calculate from the quantity and the size parameter The total amount of quinoa in the planting area; The calibration coefficient is generated according to the empirical coefficient of quinoa yield and the supply system, and the target yield of the planting area is predicted by the calibration coefficient and the total amount of quinoa. 2. The method for intelligent prediction of quinoa heading yield as claimed in claim 1, characterized in that, the quinoa particles in the planting area are scanned, and the size parameters of the current quinoa particles obtained include: The planting area is divided into a plurality of blocks, and the quinoa particles in the plurality of blocks are sequentially scanned by the collection device; The acquisition device transmits the scanned information to the upper computer, and the upper computer determines the size parameter according to the received information. 3. The method for intelligent prediction of quinoa heading yield as claimed in claim 2, wherein the scanning of the quinoa particles in the block by the collection device in turn comprises: The collection device is stabilized on one side of the corresponding block at a specific angle; The acquisition device scans the quinoa particles through pictures, videos and structured light. 4. the intelligent prediction method of quinoa heading yield as claimed in claim 2, is characterized in that, described obtaining the size parameter of current described quinoa particles comprises: The collection device obtains a plurality of characteristic points on the edge of the quinoa grain, and determines the size parameters such as the volume of the quinoa grain through the plurality of the characteristic points. 5. The intelligent prediction method for quinoa heading yield as claimed in claim 2, wherein the inferred quantity of the quinoa grains in the planting area comprises: The host computer determines the number of quinoa ears on the quinoa seedlings and the size of the quinoa ears through methods such as pictures and videos, and the host computer infers the number of the quinoa particles on the quinoa ears. , the number of the quinoa particles obtained in each of the blocks is added to finally determine the number. 6. The intelligent prediction method for quinoa heading yield as claimed in claim 5, wherein the host computer infers that the number of the quinoa particles on the quinoa ear comprises: According to the variety of quinoa, determine the number of the quinoa particles on the side of the quinoa ear facing the collecting device, and infer the total amount of the quinoa on the quinoa ear according to the size of the quinoa ear Number of quinoa grains. 7. The method for intelligent prediction of quinoa heading yield as claimed in claim 1, wherein the generation of the calibration coefficient according to the empirical coefficient of quinoa yield and the supply system comprises: Corresponding influence factors are set for each component in the supply system, the empirical coefficient includes empirical parameters for each component, and the value of each component in the supply system is compared with the value of each component in the empirical parameter. Compare the values of , and obtain the calibration coefficient from the empirical coefficient according to the influence factor. 8. The method for intelligent prediction of quinoa heading yield according to claim 7, wherein the value of each component in the supply system is compared with the value of each component in the empirical parameter, and according to the The influence factor is obtained from the empirical coefficient and the calibration coefficient includes: Predict the supply system of each interval from the current time point of quinoa to the harvest period, compare the predicted supply system with the empirical parameters, and combine the compared results with the impact factors to determine the range of each interval. The interval parameters are determined, and the calibration coefficient is finally obtained from the determined interval parameters. 9. The method for intelligently predicting the heading yield of quinoa as claimed in claim 8, wherein the result after the comparison is combined with the impact factor to determine the interval parameter in each of the described interval sections, and the determined multiple The interval parameter finally obtains the calibration coefficient including: Differentiate the value of each component in the supply system with the value of the corresponding component in the empirical parameter within the same interval, and compare the results of multiple differences with the corresponding influencing factors. Multiply, add the multiplied results to obtain the coefficient deviation, add the coefficient deviation to the empirical model to obtain the interval parameter; multiply the plurality of interval parameters to obtain the calibration coefficient . 10. The method for intelligent prediction of quinoa heading yield as claimed in claim 1, wherein the target yield predicted in the planting area by the calibration coefficient and the total amount of quinoa comprises: The target yield is obtained by multiplying the calibration coefficient by the total amount of quinoa.

Images