FR3083147A1 - SPRAY CONTROL METHOD, CORRESPONDING DEVICE AND PROGRAM - Google Patents

Abstract

The invention relates to a method for controlling the distribution of plant health product on a plant alignment, method implemented by means of a control device connected to at least one spraying device. According to the invention, such a method comprises the steps: of obtaining (10) data representative of a leaf density; and when the data representative of the leaf density exceeds a predetermined threshold: - of generation (20) of at least one command for activating at least one spray nozzle of said spray device; and - transmitting (30) said at least one spraying command to said spraying device.

Description

Spray control method, device and corresponding program

1. Domain

The invention relates to the field of plant treatments. The invention relates more particularly to the treatment of leaf hedges or of all types of plants comprising more or less dense leaf portions, such as, for example, tree-type plantations (orchards), fir stands, etc. and more generally to perennial and weeded plants. The invention thus relates to a method and to a spray control device as well as to an implementation device.

2. Prior Art

Reducing the use of phytosanitary products is at the heart of the concerns of society, public authorities and professionals. Indeed, many products appear to have negative impacts on health (toxicological and / or carcinogenic) during use or consumption and many consumers express their disapproval of the massive use made of these products. However, such products are often useful for preserving the crop, especially when it is subject to numerous pests, as is the case, for example, with vines or fruit crops. It is therefore hardly possible, given the current production methods, to do without the use of phytosanitary products. On the other hand, reducing the use of these is very beneficial. To do this, devices have been envisaged. For example, with regard to the vine, suitable diffusion devices have been proposed, in order to make the spraying more effective. This is particularly the case of the patent document AU2017202725.

Such devices, however, do not prevent an overdose of phytosanitary product. This overdose is often the result of the farmer or the professional himself, who sometimes has difficulty gauging the amount of product to be applied with regard to the plants to be treated. Furthermore, the current regulations based on doses per hectare do not take into account the different situations from one plot to another or throughout the vegetative cycle. In addition, not all plants have the same characteristics and the treatments to be applied are also different: some will focus on the base of the plant while others will directly target the leaves of it. In the latter case, it is generally important that the plant protection product reaches the maximum number of leaves. The number of leaves of a plant or group of plants is generally translated by a leaf volume. Recently, a correlation has been established between leaf density and leaf volume, especially for certain types of plants such as vines. It has been shown that one is able to infer leaf volume from leaf density. This is important proof since the quantity of phytosanitary product to be applied to a plant to guarantee its protection depends largely on the leaf volume and not on the leaf density. However, although this correlation has been established, dosages that are unsuitable for the situation and poorly adjusted sprayers remain potential sources of pollution.

Thus, although the density / volume correlation is important for the proper use of plant protection products, the fact remains that the existing techniques for applying these products are generally the same as those proposed in the past and that the quantity of product used still depends very much on the setting of the sprayer and the professional who uses this spraying device. One of the only differences is the ability to reduce the dose of phytosanitary product to be applied by calculation, that is to say before the application. By demonstrating that the leaf volume can be interpreted from the leaf surface, previous studies have made it possible to reduce the doses applied to the plantations.

On the other hand, even if the correlation between the leaf volume and the leaf surface makes it possible to reduce the doses of phytosanitary product to be applied to the plantations, the fact remains that from an environmental protection perspective, it is necessary to increase the reduction in the doses of phytosanitary product applied to plants. Indeed, a reduction in the dose on the scale of the plantation, must still be obtained. When the plantation is characterized by a disparity of plants in the plans (holes in rows of vines for example), this results in a dose actually applied to the leaves lower than that calculated since a part of the product is sprayed in a vacuum. To overcome this problem, the winegrower applies a safety factor and therefore produces a possible overdose.

There is therefore a need for a device and a method making it possible to further reduce the quantity of phytosanitary product applied to plants, and in particular leaf hedges.

3. Summary

The invention was conceived by considering these problems of the prior art. More particularly, the invention relates to a method for controlling the distribution of phytosanitary product on an alignment of plants.

More particularly, the invention relates to a method for controlling the distribution of plant health product on a plant alignment, method implemented by means of a control device connected to at least one spraying device.

According to the invention, such a method comprises the steps:

obtaining data representative of leaf density; and

When the data representative of the leaf density is not zero or exceeds a predetermined threshold:

generating at least one command for activating at least one spray nozzle of said spray device; and transmitting said at least one spray command to said spray device.

Thus, the invention makes it possible to effectively regulate the spraying of phytosanitary product as a function of the leaf density and / or of the presence of a leaf density. It is thus possible to reduce the consumption of such products as a function of the presence or not of foliage and more precisely as a function of the density of this foliage. This saves significant amounts of phytosanitary product.

According to a particular embodiment, obtaining data representative of a leaf density comprises:

a characterization of a plant environment, delivering a dense point cloud of said plant environment;

an analysis of said dense point cloud delivering said data representative of said leaf density.

According to a particular characteristic, the analysis of said point cloud delivering said data representative of said leaf density comprises:

an extraction of at least one area of interest from said cloud of dense points of said plant environment, an estimate, from said at least one area of interest, of said data representative of said leaf density.

According to a particular embodiment, the characterization of a plant environment includes:

a movement of an instrumented carrier carrying a contactless sensor delivering a plurality of samples of geometric data in three dimensions;

a reconstruction of said plant environment from said plurality of three-dimensional geometric data samples, delivering the dense point cloud of said plant environment.

According to a particular characteristic, the extraction of at least one area of interest from said dense point cloud of said plant environment comprises:

deleting, from said dense point cloud, points located outside a predetermined volume, delivering a delimited point cloud;

an estimate, within the defined point cloud, of a point density, delivering said leaf density.

According to another aspect, the invention also relates to a device for controlling the distribution of plant health product on a plant alignment, said device being connected to at least one spraying device. Such a device comprises means:

obtaining data representative of leaf density; and generating at least one command for activating at least one spray nozzle of said spray device; and transmitting said at least one spraying command to said spraying device, said generation and transmission means being activated when the data representative of the leaf density exceeds a predetermined threshold.

According to a particular characteristic, said means for obtaining data representative of a leaf density comprise means for capturing said plant environment, said means for capturing comprising a sensor of three-dimensional geometry.

According to a particular embodiment, the three-dimensional geometry sensor comprises a LiDAR.

According to another aspect, the invention also relates to a system for distributing plant health product on an alignment of plants. Such a system includes:

at least one mobile carrier;

a phytosanitary product distribution control device as described above, located on said at least one mobile carrier;

a phytosanitary product spraying system linked to said phytosanitary product distribution control device, located on said at least one mobile carrier; and means allowing the displacement of said at least one mobile carrier.

According to a preferred implementation, the different steps of the methods according to the invention are implemented by one or more software or computer programs, comprising software instructions intended to be executed by a data processor according to the invention and being designed to order the execution of the different process steps.

Consequently, the invention also relates to a program capable of being executed by a computer or by a data processor, this program comprising instructions for controlling the execution of the steps of a method as mentioned above.

This program can use any programming language, and be in the form of source code, object code, or intermediate code between source code and object code, such as in a partially compiled form, or in any other desirable form.

The invention also relates to an information medium readable by a data processor, and comprising instructions of a program as mentioned above.

The information medium can be any entity or device capable of storing the program. For example, the support may include a storage means, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or else a magnetic recording means or a hard disk or a memory card.

On the other hand, the information medium can be a transmissible medium such as an electrical or optical signal, which can be routed via an electrical or optical cable, by radio or by other means. The program according to the invention can in particular be downloaded from a network of the Internet type.

Alternatively, the information medium can be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the process in question.

According to one embodiment, the invention is implemented by means of software and / or hardware components. In this perspective, the term module can correspond in this document as well to a software component, as to a hardware component or to a set of hardware and software components.

A software component corresponds to one or more computer programs, one or more subroutines of a program, or more generally to any element of a program or of software capable of implementing a function or a set of functions, as described below for the module concerned. Such a software component is executed by a data processor of a physical entity (terminal, server, gateway, router, etc.) and is capable of accessing the hardware resources of this physical entity (memories, recording media, bus communication, electronic input / output cards, user interfaces, etc.).

In the same way, a hardware component corresponds to any element of a hardware assembly (or hardware) capable of implementing a function or a set of functions, according to what is described below for the module concerned. It can be a programmable hardware component or with an integrated processor for the execution of software, for example an integrated circuit, a smart card, a memory card, an electronic card for the execution of firmware ( firmware), etc.

4. Drawings

Other characteristics and advantages of the invention will appear more clearly on reading the following description of a preferred embodiment, given by way of simple illustrative and nonlimiting example, and of the appended drawings, among which:

FIG. 1 presents in general the principle of controlling the distribution of phytosanitary product according to the invention;

FIG. 2 briefly describes a device capable of implementing the method presented in FIG. 1;

FIG. 3 describes a method of obtaining a point cloud from a LiDAR in one embodiment;

Figure 4 is a representation of a dense point cloud of said plant environment;

FIG. 5 shows a delimited point cloud obtained from a dense point cloud according to a particular embodiment of the invention;

FIG. 6 presents the principle of processing the data of the dense point cloud from sliding windows;

FIG. 7 briefly describes the architecture of a control device according to one embodiment.

5. Description

5.1. Reminders of general characteristics

The general principle of the invention consists in differentiating the spraying of phytosanitary product as a function of the leaf density (which we have seen that it is linked to the leaf volume) of a set of plants aligned in the form for example of a hedge. The invention, in its general principle, is applicable to all types of plants, in the form of an alignment, whether they are hedges, vines, or even fruit trees planted in an orchard, but also to weeded plants. The configuration modes of the collection system and the spraying system are directly linked to the plants to be treated.

More particularly, there is described a method and a device for controlling the distribution of phytosanitary product on an alignment of plants. This method and this device are advantageously implemented on a carrier, moving along the alignment of plants so as to allow the application of the phytosanitary product continuously, throughout the movement of the carrier. The method comprises the following steps, described in relation to FIG. 1:

obtaining (10) data representative of a leaf density of the plant alignment or of a portion of the plant alignment; and when the data representative of the leaf density is not zero or when it exceeds a predetermined threshold:

generation (20) of at least one activation command of at least one spray nozzle of said spray device; and transmitting (30) said at least one spray command to said spray device.

The method therefore makes it possible to correlate the distribution of plant protection product to the presence of plants in a sufficient density (and therefore a volume) of foliar (that is to say above a predetermined threshold) or then all or nothing when only presence is taken into account. Naturally, obtaining a data representative of a leaf density of the plant alignment or of a portion of the plant alignment can be implemented according to several different embodiments. However, it requires the use of sensors, which, depending on their nature, will make it possible to obtain data having different characteristics. Furthermore, obtaining data representative of leaf density can be implemented in real time, that is to say as and when a carrier on which the control device is installed is moved; obtaining data representative of leaf density can be implemented prior to the passage of the carrier, for example by obtaining data using a capture system which operates independently of the control device, the capture system having '' a positioning mechanism which allows precise localization of leaf densities.

In any event, once the leaf densities have been obtained (in real time or not), these are used to control the activation of the spraying by generating suitable commands, in particular for example in terms of height relative to the nozzle location on the side of the sprayer or in terms of location. The transmission of commands is carried out when the nozzle (s) concerned is at the location where the measured leaf density requires spraying. This saves a significant amount of product potentially harmful to the environment (when used in excess) and which also have a relatively high cost (financially, in terms of product image, in terms of possible subsidies lost or reduced, etc.), while ensuring a minimum dose sprayed.

Obtaining data representative of a leaf density comprises, in one embodiment, a characterization (10-1) of a plant environment, delivering a dense point cloud of said plant environment and an analysis (10-2 ) of said dense point cloud delivering said data representative of said leaf density. The point cloud is obtained through a sensor which has the capacity to provide such data. More particularly, the dense point cloud is a three-dimensional point cloud, in which the points represent the presence of an element, plant or not, which is characterized. As explained in relation to the embodiment described below, the analysis according to the characterization consists in particular in delimiting an area of interest representative of the leaf density, and it comprises in particular:

an extraction (10-21) of at least one zone of interest from said cloud of dense points of said plant environment; the area of interest extracted is for example a volume (included in the point cloud), volume which is configured in advance according to the nature of the plants to be treated (hedge, vine, orchard).

an estimate (10-22), from said at least one area of interest, of said data representative of said leaf density: the points (of the point cloud) which are included in this volume serve to determine the leaf density; for example, it may be decided that the density of points is directly linked to the density of leaves (direct method); other methods of estimation from the point density of the point cloud can also be implemented.

The density being obtained for this area of interest, it is then possible to control the spraying as a function of the value of this density, in the area of interest considered.

Although the process which is the subject of the invention has been described in the context of the application of plant protection product to vines, it is understood that this process can also be implemented in the context of the application of other substances on other types of plants, starting with the application of an aqueous spray, for example in the context of hydration operations. The aqueous spraying can be carried out either on the leaves (if this is useful), or at the base of the plant, depending on the leaf density of the latter (considering for example that a plant having a high leaf density will need more water).

5.2. Description of an embodiment

In this embodiment described in connection with FIG. 2, a spray control system (SysC) is presented, implementing the spray control method presented above. More specifically, in this embodiment, a spray control device (DispC) is implemented on a carrier (Port), which moves along rows of vines. In this embodiment, it is an autonomous carrier. However, this characteristic of autonomy is not essential for the implementation of the spraying process and the carrier can quite be towed by a motorized agricultural machine and driven by a user; or else the carrier can be the agricultural machine itself. Anyway, the spraying control device is mounted on a carrier and comprises the means necessary for the implementation of the invention and in particular, means for characterizing the environment in three dimensions, comprising one or more several sensors (CapT), the nature of which is specified below. These sensors are used to obtain "raw" information from the plant environment (foliar). This raw information is combined and segmented (free of noise, artefacts) to deliver an effective leaf density, which takes the form of a percentage of plan or volume occupied by vine leaves. The means making it possible to segment the raw information and to characterize it (for example in the form of a percentage) are typically in the form of a calculation module comprising a processor, a memory (s) and entries exits. Depending on the percentage (therefore the leaf density of the area observed by the sensors), the device then generates control signals intended for nozzles (or solenoid valves coupled to nozzles) in order to activate spraying in a binary manner or proportionally (depending on the leaf density). The spraying system (SPuI) as such (for example the phytosanitary product reservoir, the lines for driving the phytosanitary product towards the spray nozzles (BU), the suspension arms, the spraying control means (solenoid valves ), etc.) is for example arranged on an additional carrier or on the same carrier (Port) as that which has the spraying control device.

In this embodiment, the carrier (and therefore the device) moves in the row of vines and calculates, in real time and continuously (as and when moving) the leaf density in this row (either of only one side, ie on both sides). This density is used to activate the spray nozzles. The calculation of leaf density is carried out by grouping (merges, combinations, readjustments) of several successive data captures, captures carried out by means of characterization of the environment in three dimensions. The registration is carried out on the basis of the displacement estimate described below. Insofar as the transformation between the reference mark of the sensor and the reference mark of the wearer is known, it is possible to readjust the different successive captures. From the movement of the carrier between two captures, the registration is carried out by transforming the successive captures.

5.2.1. 3D characterization

As explained above, the 3D characterization consists, in this embodiment, of having a three-dimensional representation of the environment within which the wearer moves. More particularly, it is a matter of having a representation of the vegetable hedge located on at least one of the two sides of the carrier.

In this embodiment, the three-dimensional acquisition technology implemented is based on the use of a LiDAR. Given the existing constraints and current limitations of alternative techniques, this technology has several advantages, including the possibility of directly analyzing a point cloud (rather than having to derive this point cloud from a set of images or data obtained through other technology). Furthermore, the inventors have found that this technology is proven in an outdoor environment and proves to be robust in relation to variable conditions (variations in meteorology, brightness, amplitude of vibrations, etc.). Although the technique used can be sensitive to vibrations, the impact of these is limited, in particular due to the relatively coarse measurement precision / resolution, as explained below. The weight of the carrier is also a limiting factor for the impacts of vibrations. In addition, this technology delivers measurements in real time - on the order of a few tens of milliseconds per profile capture.

In order to allow a satisfactory characterization, the sensor is physically mounted on the carrier so that the capture is carried out by aiming at the ground. The sensor is in fact directed towards the ground in order to obtain a vertical section (or a profile) of the row of vines at a given instant. By moving the LiDAR sensor inside a vine row using the carrier, and by resetting the captured profiles according to the displacement, part of the row is characterized: this part of the row is then composed of several cuts (Cl, ..., Cn) as illustrated in Figure 3.

One of the problems to be solved consists in constructing, a satisfactory representation of one or part of the set of foliage, from these sections. The use of a very precise positioning device cannot be envisaged, in particular for cost reasons, and is not essential for the implementation of the porposed technique. Thus, the construction of a point cloud representative of all the foliage is carried out by estimating the displacement of the carrier on which the capture device (the LIDAR in this case) is positioned. Generally, the movement of the carrier is used to construct the sets of points from the different sections. Several displacement measurement mechanisms can be implemented to solve this problem. In this embodiment, the movement of the carrier is estimated from the odometry of the wheels and from a gyroscope. The extended odometry (path) is therefore derived from the individual measurement of sensors (for example present in each wheel, but one or two sensors may suffice) combined with the absolute orientation obtained from a gyroscope. This combined measurement makes it possible to reconstruct the overall movement of the wearer from a specific model. Starting from a known initial position and integrating the measured displacements and orientations, it is thus possible to calculate at each instant the current pose of the carrier (position and orientation) and therefore to estimate the absolute path traveled by the carrier in the “world” mark (ie from the known position). According to the present technique, it will be noted that although the proposed method may include some limitations, in particular due to sliding, slope (and therefore estimation of the absolute path over a long distance), it remains valid over short distances , and therefore more particularly on the distance separating the sensor (for example the LiDAR) from the spray nozzles. In other words, it is necessary and sufficient that the distance is correctly estimated between the capture of the density and the spraying in order to fulfill the objective of intelligent spraying.

Among other possibilities of displacement measurement, RGB-D SLAM techniques which consist of a displacement of a depth sensor and a dense 3D reconstruction without resorting to the wearer's odometry can also be implemented. The advantages are to be able to use sensors at a lower cost (Time Of Flight Cameras) and to be independent of the wearer's instrumentation (the odometry is intrinsically calculated by the method). The drawbacks mainly lie in the large computing capacity necessary for dense 3D reconstruction and in the lower spatial resolution and precision obtained compared to LiDAR. On the other hand, when the intrinsic leaf density of the plant environment is known (for example on a hedge made up of plants of which it is known that the leaf density is constant at the locations where there are leaves), the consequences of this less precision are limited.

Consequently, the choice of a type of sensor compared to a set of available sensors essentially depends on the operational implementation conditions and the quality expected in the estimate. The choice of LiDAR, made by the inventors, gives good results, but other sensors, which are less expensive, can also be used when the precision of the estimation of the leaf density is not a determining criterion (for example example because this density is more or less constant).

5.2.2. Pretreatments and Analysis

Data from the sensor (s) is obtained in real time. The construction of the point cloud of the environment located around the carrier is achieved by accumulating the points from successive captures, as and when the movement. An example of a point cloud is presented in relation to Figure 4.

The spatial resolution (distance separating each point captured) is defined according to two independent parameters:

The vertical and horizontal resolution is determined by the angular resolution of the sensor as well as by the distance separating the hedge from the sensor.

The longitudinal resolution, defined along the axis of the row (or the axis of displacement), which is determined by the speed of advance of the carrier.

The data is then segmented in order to keep only the useful data of the leaf hedge. Thus, a data preprocessing method is implemented. It consists of creating a bounding box around the sensor (areas of interest), in order to remove the soil as well as any other element outside of this box (rows of vines other than the two rows adjacent to the carrier, or other data not relevant outside the important space). The results obtained by applying the pretreatments to the data illustrated in Figure 4 are shown in Figure 5.

Thus, as illustrated in FIG. 5, there is, after the pretreatment, a cloud of points delimited at a determined volume. In this case, there is a point cloud representative of the two rows adjacent to the carrier in the example of FIG. 4. Of course, when a single row is concerned, the delimited point cloud only relates to to this single rank.

The leaf density is estimated from this delimited point cloud. In this embodiment, it is estimated as a function of the density of points captured in a given volume. It is indeed considered that the density of the foliage is directly proportional to the density of points measured in a given volume. Other indicators used to refine the estimate have been developed in the literature dealing with agronomic characteristics (such as leaf volume or LAD). They are not used in this embodiment. In other embodiments, which would have, for example, additional computational resources or different capture methods, these indicators could be derived from the measurements carried out and lead to the obtaining of leaf density just as in this first embodiment.

In this embodiment, the approach chosen for the analysis of the leaf density consists of an implementation of a sliding window mechanism for analyzing the captured 3D points. The operating cycle of the sliding window system is shown in Figure 6. It illustrates the operating principle of a sliding window. In the figure, an exemplary embodiment is given in detail by a sliding window (F1, F2, Fn) and a data set composed of twelve values. In this example, the window encompasses several sets of data at the same time, for example at the start of the first eight, then moving one step to the right to encompass the next eight data, and so on iteratively. At each movement, a calculation is performed with the values included.

Leaf density is estimated using sliding windows. This estimate is made according to several geometric parameters:

the minimum height (z _min ) and the maximum height (z _max ) make it possible to set a range of values to be analyzed on the height. For example, it is not useful to keep the ground or the points located more than two meters when a hedge (for example of a vine) typically measures 1.20 meters in height above the ground. the number of blocks (p) over the height (difference (z _max ) minus (z _min )) and the maximum height makes it possible to control the vertical resolution: the number of blocks is directly correlated to the number of nozzles; if for example there are six nozzles distributed over the maximum height, the enclosed window is divided (in height) into six blocks, each block for which a leaf density is calculated corresponding to a nozzle capable of being activated for this block.

Thus, one obtains a set of “encompassing” cubes resulting from a spatial sub-sampling of the analyzed point cloud.

The density value is estimated by counting the number of points contained in the sliding window. The data is conditioned in the form of a percentage of filling of the volume considered. This conditioning makes it possible to remain generic with respect to the spatial resolution of the 3D sensor used. Anyway, the leaf density is determined by sliding the window (and therefore the progression of the carrier). The block density data can advantageously be kept (either in the carrier or elsewhere) for later recording purposes (for example to create an overall leaf density map. This overall map can be compared throughout the growth of the leaf hedge, for example to establish growth statistics or growth trends.

Leaf density can be estimated instantly. The use of sliding analysis windows allows the density of the blocks to be calculated at each time sampling, therefore for each profile and in real time. The density is thus delivered for each profile. But it is not only estimated on the current profile but on a sliding window in order to integrate several values and thus reduce the noise generated by the measurement. Insofar as there is a high spatial resolution, it is therefore possible to apply an integration by sliding windows. The width of the windows depends directly on the spatial resolution (defined above): the longitudinal resolution, defined along the axis of the row (or of the axis of displacement), which is determined by the speed of advance of the wearer.

In the case of a static analysis, several measurements are made and a value is extracted from statistics on the measurements.

5.2.3. Generation of Command Signals

The leaf density values estimated in the previous module are converted into a set of associated control signals. The generated signals can be of types:

binary for an All-Nothing control (activation / deactivation of the spray nozzles) analog (modulation of the spray nozzle flow rate, allowing to modify the volume (the quantity) of sprayed phytosanitary product).

All or Nothing Order

Binary values are calculated by applying a minimum density threshold to the leaf density value. This threshold is evaluated according to the application (from 5% to 30%) for example after analysis of the preliminary results. This threshold thus makes it possible to affect:

a value of 0 (Nothing) in areas with low or no leaf density (located below a predetermined threshold value, for example from 5 to 30%);

a value of 1 (All) in areas with significant leaf density (located below a predetermined threshold value, for example above 5 to 30%).

The generation of On / Off control signals is carried out using a power relay control card. (The card is controlled from the on-board computer on which the foliar density calculation algorithm is installed) which transmits an electrical control signal to the relay which activates for example a solenoid valve connected to a hose of the spraying system, this solenoid valve enabling allow the phytosanitary product to flow to the spray nozzle. Depending on the embodiments, several solenoid valves can be used, for example one per nozzle. Essentially, the dimensioning of the number of nozzles and solenoid valves depends on the operational implementation conditions.

Analog control

An alternative to this embodiment is to generate analog control signals. The analog values here are directly derived from the values of leaf density expressed as a percentage.

The generation of analog control signals can be carried out using a card with analog outputs or with PWM outputs (The card is controlled from the on-board computer on which the foliar density calculation algorithm is installed). In this case, a variable flow solenoid valve can be used on the spraying system to reflect the variation in the amount of phytosanitary product to be applied as a function of the variation in leaf density. In another embodiment, a variable flow nozzle can be used in place of the variable flow solenoid valve.

5.2.4. Controlled Spray

The control signals generated as a function of the leaf density are used to control a set of controllable nozzles for spraying plant protection product. The system allows a command to open and close the nozzles (All or Nothing command).

To control each nozzle by an All-or-Nothing command, it is possible to implement a specific system for supplying the liquid under constant pressure, designed by the inventors. Each nozzle is then continuously supplied and a complementary system of solenoid valves is used to allow the product to return to the tank as soon as the spraying is not activated.

In a specific embodiment, a combination of two two-way solenoid valves is used to control each nozzle. The embodiment chosen, depending on the height of the plants, consists for example of an integration of a set of four nozzles distributed in pairs on each side on a vertical ramp. This thus makes it possible to have a differentiated spraying control on the top and the bottom of the boom as well as on the left and right sides of the carrier. The system implements a spray jet spray using turbulence nozzles.

An alternative to this embodiment, as explained above, consists in modulating the flow rate of the nozzles from analog control signals (analog or digital electrical signals of PWM type).

5.3. Other features and benefits

The main components of a plant protection product distribution control device are described in relation to FIG. 7, in a particular embodiment. Such a device comprises means for determining a leaf density of an alignment of plants according to measurements carried out by a capture device. It also includes means for generating commands to an electro-controlled spraying device as a function of the estimated leaf density.

For example, the phytosanitary product distribution control device according to the proposed technique comprises a memory 71 consisting of a buffer memory (M), a processing unit 72, equipped for example with a microprocessor (μΡ), and controlled by the computer program (Pg) 73, implementing the data processing method according to the invention.

On initialization, the code instructions of the computer program 73 are for example loaded into a memory before being executed by the processor of the processing unit 72. The processing unit 72 receives as input (E) for example data corresponding to the values from the environmental sensors (from the alignment of plants). These values change over time, depending on the movement of the device on a carrier. The microprocessor of the processing unit 72 performs the steps of the control method for obtaining the leaf density, in real time or in deferred time, according to the instructions of the computer program 73, to determine at least one command and finally transmit it at the output (T) to a spraying device.

More particularly, the microprocessor of the processing unit 72 is capable of, or includes means:

obtaining data representative of a leaf density from external sensors; and generating at least one command for activating at least one spray nozzle of said spray device; and transmitting said at least one spray command to said spray device.

Claims (10)

1. Method for controlling the distribution of plant health product on a plant alignment, method implemented by means of a control device connected to at least one spraying device, method characterized in that it comprises the steps: obtaining (10) data representative of a leaf density; and when the data representative of the leaf density is not zero or exceeds a predetermined threshold: generating (20) at least one activation command for at least one spray nozzle of said spray device; and transmitting (30) said at least one spray command to said spray device. 2. Control method according to claim 1, characterized in that obtaining data representative of a leaf density comprises: a characterization (10-1) of a plant environment, delivering a dense point cloud of said plant environment; an analysis (10-2) of said dense point cloud delivering said data representative of said leaf density. 3. Control method according to claim 2, characterized in that the analysis of said point cloud delivering said data representative of said leaf density comprises: an extraction (10-21) of at least one area of interest from said dense point cloud of said plant environment, an estimate (10-22), from said at least one area of interest, of said data representative of said leaf density. 4. Control method according to claim 2, characterized in that the characterization of a plant environment comprises: a displacement of an instrumented carrier carrying a contactless sensor delivering a plurality of samples of geometric data in three dimensions; a reconstruction of said plant environment from said plurality of three-dimensional geometric data samples, delivering the dense point cloud of said plant environment. 5. Control method according to claim 3, characterized in that the extraction of at least one zone of interest from said dense point cloud of said plant environment comprises: deleting, from said dense point cloud, points located outside a predetermined volume, delivering a delimited point cloud; an estimate, within the defined point cloud, of a point density, delivering said leaf density. 6. Device for controlling the distribution of plant protection product on a plant alignment, said device being connected to at least one spraying device, device characterized in that it comprises means: for obtaining data representative of a leaf density; and generating at least one command for activating at least one spray nozzle of said spray device; and transmitting said at least one spraying command to said spraying device, said generation and transmission means being activated when the data representative of the leaf density exceeds a predetermined threshold. 7. Device for controlling the distribution of plant health product on a plant alignment, according to claim 6, characterized in that said means for obtaining data representative of a leaf density comprise means for capturing said plant environment, said means capture means comprising a three-dimensional geometry sensor. 8. Device for controlling the distribution of plant protection product on a plant alignment, according to claim 7, characterized in that the three-dimensional geometry sensor comprises a LiDAR. 9. Plant protection product distribution system on an alignment of plants, system characterized in that it comprises: at least one mobile carrier; a plant protection product distribution control device according to any one of claims 6 and 7, located on said at least one mobile carrier; a phytosanitary product spraying system linked to said phytosanitary product distribution control device, located on said at least one mobile carrier; and means allowing the displacement of said at least one mobile carrier. 10. Computer program product downloadable from a communication network and / or stored on a computer-readable medium and / or executable by a microprocessor, characterized in that it comprises program code instructions for the execution of a spray control method according to any of claims 1 to 5 when executed by a processor.