SUBJECT TO ORGANIC FARMING

DESCRIPTION

The present invention relates to a field monitoring process used to determine whether the field is subject to organic farming, as well as to a system that implements such a process.

Organic farming is growing steadily, with a 400% expansion in the production area from 1999 to 2014. Such an increase is due to the growing demand of organic products that, compared with conventional products, are perceived by consumers as healthy products, with a low environmental impact and zero presence of synthetic chemical/mineral products.

In order to certify that their agricultural products are organic and sell them with the ORGANIC label, farmers must comply with international standards that contain well-defined rules.

The annual certification of the organic agricultural products comprises:

1 . An on-site inspection of the fields declared to be organic

2. Analysis of residues of agricultural-pharmaceutical products of batches delivered for a residual analysis

3. Drafting of a field book

4. Analysis of the invoices issued by the farm.

Such a certification procedure is very time-consuming and expensive, especially for large farms that also have plots in remote and difficult-to-access areas.

In addition, consumers have growing doubts as to whether the products with the ORGANIC label are actually healthy products grown in fields worked with cultivation procedures that have no impact on the environment and contain no chemicals.

US10609860B1 describes a system for estimating nitrogen content in plants, wherein hyperspectral and/or multispectral remote sensing images are automatically analyzed by a nitrogen analysis subsystem to estimate the value of the nitrogen variables in the crops of the images. Such a system provides for an artificial intelligence algorithm that trains a mapping function to estimate the values of the nitrogen variables for a new remote sensing image that is not included in the training dataset. Said nitrogen values can be used to determine an optimal amount of fertilizer to be added to a crop field in order to promote plant growth.

The purpose of the present invention is to eliminate the drawbacks of the prior art by providing a field monitoring process and system to automatically determine whether the field is used for organic or non-organic farming and whether the farmer declared to be organic is performing practices that are prohibited by the international standard procedures of organic farming.

Another purpose is to provide such a field monitoring process and system that are likely to reduce the cost and time of the organic certification of the field that have to be borne by the certification bodies and the farmers.

Another purpose is to provide such a field monitoring process and system that are efficient, reliable, practical and accurate.

These purposes are achieved in accordance with the invention with the features of the appended independent claims.

Advantageous achievements of the invention appear from the dependent claims.

Further features of the invention will appear clearer from the following detailed description referring to a purely illustrative and therefore non-limiting embodiment illustrated in the appended drawings, wherein

Fig. 1 is a block diagram schematically illustrating the field monitoring system according to the invention;

Fig. 2 is a schematic view of the loT sensors of the system of Fig. 1 ,

Fig. 3 a boxplot graph illustrating a variation in the vegetation index in a fertilized field (conventional farming) and a non-fertilized field (organic farming) over a period of about 2 months in 3 different years;

Fig. 4 is a histogram graph illustrating the variation in the vegetation index of the graph of Fig. 3;

Fig. 5 is a Time-Series graph illustrating a variation in the vegetation index of the graph of Fig. 3;

Fig. 6 is a block diagram illustrating a training phase of the algorithm of the monitoring system according to the invention;

Fig. 7 is a block diagram illustrating a classification phase performed by the trained algorithm of the monitoring system according to the invention.

With the aid of the Figures, a field monitoring system according to the invention, which is comprehensively denoted with reference numeral 100, is described. Fig. 1 diagrammatically illustrates a field (200) that is to be monitored by the system (100) to determine whether the field (200) is used for organic farming.

The system (100) comprises:

- a data processing unit (1 ),

- pedological sensors (2) arranged in the field (200) to detect pedological data (D1 ) of the field,

- climatic sensors (3) arranged above the field (200) to detect climatic data (D2) indicative of the climatic conditions of the field,

- a GPS (6) installed in the field (200) to detect a geographical position (Pg) of the field,

- an image processing unit (9), and

- image acquisition means (7) suitable for capturing images (I) of the field (200).

The image acquisition means (7) are connected to the image processing unit (8) to send the images (Is, Id) of the field to the image processing unit (8).

The image processing unit (8) is configured to receive the images (Is, Id) of the field and calculate at least one vegetation index (V) of the field. The image processing unit (8) is connected to the data processing unit (1 ) to send the vegetation index (V) to the data processing unit (1 ).

The pedological sensors (2) and the climatic sensors (3) are connected to the data processing unit (1 ) to send the detected pedological data (D1 ) and the climatic data (D2) to the data processing unit (1 ).

Advantageously, the system (100) may also comprise portable sensors (4) placed in the field (200) as needed to detect crop data (D3) that is not detected by the pedological sensors and the climatic sensors.

Advantageously, the data processing unit (1 ) can be a server on an Internet cloud, based on artificial intelligence, as will be explained in detail.

Advantageously, the pedological sensors (2) and the climatic sensors (3) can be loT (Internet of Things) sensors. In such a case, the pedological sensors (2) and the climatic sensors (3) can be directly connected to the Internet or they can be wirelessly connected to a hub (5) placed near the loT sensors, which connects to the Internet to send the pedological data and the climatic data (D1 , D2) detected by the loT sensors to the data processing unit (1 ).

Referring to Fig. 2, the pedological sensors (2) may include one or more of the following sensors: - a first temperature sensor (2a) placed in the soil of the field (200) to detect a temperature (T1 ) of the soil of the field;

- a first humidity sensor (2b) placed in the soil of the field (200) to detect a relative humidity (U1) of the soil of the field;

- a first hydrometer (2c) placed in the soil of the field (200) to detect a water content ( Ig 1 ) of the soil of the field;

- a nitrogen sensor (2d) placed in the soil of the field (200) to detect a nitrogen (N) content of the soil of the field;

- a phosphorus sensor (2e) placed in the soil of the field (200) to detect a phosphorus (P) content in the soil of the field; and

- a potassium sensor (2f) placed in the soil of the field (200) to detect a potassium (K) content in the soil of the field.

The climatic sensors (3) may include one or more of the following sensors:

- a second temperature sensor (3a) placed above the field (200) to detect a temperature (T2) of the air above the field;

- a second humidity sensor (3b) arranged above the field (200) to detect a relative humidity (U2) of the air above the field;

- a second hydrometer (3c) placed on leaves of the crop of the field (200) to detect a leaf wetness (Ig2) of the leaves of the crop of the field; and

- an anemometer (3f) placed above the field (200) to detect a wind speed (Vf) and a wind direction (Vf1).

Obviously, two or more of the above-mentioned climatic and pedological sensors can be integrated into a single device, such as a climatic pedological cell, or separate autonomous sensors can be used.

The GPS system (6) is connected to the data processing unit (1). Therefore, the GPS system (6) detects the geographical position (P) of the field and sends it to the data processing unit (1). It should be considered that, being loT sensors, the pedological and climatic sensors (2, 3) generally have their own GPS system. Therefore, the GPS system (6) can be integrated in the pedological and climatic sensors (2, 3).

The image acquisition means (7) may include:

- a satellite image acquisition system (70) in which satellite images are stored, including the satellite images (Is) of the field (200) being monitored; and/or

- a drone image acquisition system (71) in which cameras installed in drones acquire images (Id) of the field (200) being monitored. The satellite image acquisition system (70) and the drone image acquisition system (71 ) are connected to the GPS system (6) to select the images related to the geographic location (Pg) of the field to be monitored.

In the case of satellite image acquisition systems (70), the satellite images are multispectral or hyperspectral or thermal or synthetic radar aperture images. For example, a type of satellite image cab be derived from satellite acquisition systems such as Copernicus (a free service of the European Union) or commercial services.

In the case of a drone imaging system (71 ), drones are equipped with cameras suitable for acquiring images of the field to be monitored.

In any case, based on the images (Is, Id) detected by the image detection means (7), the image processing unit (8) is able to calculate at least a vegetation index (V) of the crop of the field (200) which is sent to the data processing unit (1 ).

Preferably, the NDVI (Normalized Difference Vegetation Index) or the NDRE (Normalized Difference Red Edge) vegetation index is calculated. The NDVI describes the level of vigor of the crop and is calculated as the ratio of the difference to the sum of the reflected radiation in the near infrared (NIR) and red (RED). The NDRE vegetation index is similar to the NDVI, but uses the reflected radiation in the red-edge instead of red.

With reference to Fig. 3, a boxplot graph is illustrated, which shows values of the NDRE vegetation index in a fertilized field and in a non-fertilized field.

In an initial situation as of 28 March 2018, both fields start from the same value of the vegetation index. With the passing of time, the vegetation index of the fertilized field increases compared to the vegetation index of the non-fertilized field.

Also the graphs in Figs. 4 and 5 clearly show that in a three-month period, in three different years, the value of the vegetation index of the fertilized field is higher than the value of the vegetation index of the non-fertilized field.

Such information on the vegetation index of a fertilized field and of a nonfertilized field is a discriminating factor in determining whether a field is subject to organic farming or conventional farming.

Obviously, the values of the vegetation index (V) must be closely related to the pedological data (D1 ), to the climatic data (D2) and to the geographic location (Pg) of the field.

With reference to Fig. 6, the data processing unit (1 ) comprises an artificial intelligence algorithm (10). The artificial intelligence algorithm (10) must be trained with a calibration and validation process. The calibration and validation process uses as reference at least one field (200A) on which it is certain that organic farming activities are being conducted and at least one field (200B) on which conventional farming activities are being conducted.

The fields (200A, 200B) are monitored continuously with the system (100), calculating the respective vegetation indices (Va, Vb) correlated with the respective pedological data (D1 a, D1 b), the climatic data (D2a, D2b), the geographic location (Pga, Pgb) and possibly also with the respective crop data (D3a, D3b) detected by the portable sensors.

Obviously, since the pedological data (D1 a, D1 b) and the climatic data (D2a, D2b) are continuously detected by the loT sensors, the data processing means (1 ) are configured to calculate a time average of such data, for example over the period of one day with a given sampling frequency. A substantially similar average can also be done with the vegetation indices (Va, Vb) and with the crop data (D3a, D3b).

In such a way, based on the data received by the algorithm (10), the following are defined: i) a first reference sample related to the vegetation index correlated with the pedological data, the climatic data and the geographic location data of the organic field (200A); and ii) a first reference sample related to the vegetation index correlated with the pedological data, the climatic data and the geographical position data of the non- organic field (200B).

The first reference sample and the second reference sample serve as terms of comparison for the subsequent classifications of the monitored fields.

At the end of the training phase, the algorithm (10) collected all the data of the field (200A) with organic farming and all the data of the field (200B) with non-organic farming. The collected data are sample data that will be used as comparison data with the data acquired during the classification phase.

With reference to Fig. 7, in the classification phase, the algorithm (10) performs a classification of each field (200) monitored by the system (100) and classifies it as organic or non-organic field.

In this stage, the algorithm (10) continuously receives the pedological data (D1 ), the climatic data (D2) and the vegetation index (V) for the field (200) to be monitored, and compares them with the first reference sample and the second reference sample obtained in the training phase. Based on such a comparison, the algorithm classifies the field (200) as organic or non-organic. Obviously, it may occur that a part of a field is organic and another part of a field is non-organic. In such a case the algorithm (10) can make a classification on a percentage basis and generate an output such as the following:

- 60% organic field

- 40% conventional field.

The data (D1 , D2, D3) and the vegetation index (V) received by the data processing unit (1 ) can be displayed by an application that will enable the farmer to analyze the data in order to ensure a proper agronomic management of the plots by performing precision farming practices.

In addition, the data processing unit (1 ) will send the data (D1 , D2, D3), the vegetation index (V), and the result (organic or non-organic) in an output of the system that can communicate with numerous certification bodies.

The output of the system:

- sends the result of the artificial intelligence algorithm directly to certification bodies, which can use the result to certify the agricultural farmer;

- generates a report that can be submitted by the farm to the authorities to obtain an organic certification;

- after indicating the site being analyzed and the data acquired by the operator, API Service - that is to say the certification body - can send that data to the system, which then outputs the response indicating whether the field is organic on non-organic.

Equivalent variations and modifications may be made to the present embodiment of the invention, within the reach of an expert of the field, but still within the scope of the invention as expressed by the appended claims.