WO2023275698A1

WO2023275698A1 - Sensor for the detection of cultivated plants and parts of the same

Info

Publication number: WO2023275698A1
Application number: PCT/IB2022/055927
Authority: WO
Inventors: Nicola Baldo; Ulrich Hausmann
Original assignee: Spray Logics S.R.L.
Priority date: 2021-06-28
Filing date: 2022-06-27
Publication date: 2023-01-05
Also published as: IT202100016892A1

Abstract

The present invention finds its application, in general, in the agricultural sector. In particular, the solution concerns a transportable sensor device, which is essential for optimizing and automating the treatment processes that must be carried out on fruit plants in general. Ultimately, the sensor device according to the teachings of the invention can be used to acquire a two-dimensional map of a row of plants, without going through the acquisition of an image in the visible field which would involve the use of optical sensors, whose use presents various problems; instead reconstructing equivalent images (i.e., maps describing a row of plants seen from its side) which are very significant from the point of view of their information content. In fact, such equivalent images acquired by the device indicated in the invention do not require to be reprocessed through recognition programs, but directly contain the information that is needed in many types of applications. The invention is based on an innovative use of infrared radiation active sensor technology, which however requires the design and implementation of a new type of sensor device, built with specific new features, such as the use of very directive radiation, in narrowband and on a plurality of carriers. Furthermore, also the geometric structure, and the shape, of the sensor device itself must have some particular features, which make it suitable for performing a sort of scanning of entire rows of plants; everything in a simple way and without the need to resort to complex handling mechanisms.

Description

TITLE: SENSOR FOR THE DETECTION OF CULTIVATED PLANTS AND

PARTS OF THE SAME

DESCRIPTION

Technical Field of the Invention The present invention finds its application, in general, in the agricultural sector. In particular, the solution indicates a transportable sensor device, which is essential for optimizing and automating the treatment processes that must be carried out on fruit plants in general.

State of the Art Vineyards and orchards represent a very particular agricultural sector, not one of the simplest to automate.

The modern layouts of orchards or vineyards are conceived in such a way as to arrange the plants in rows (in some cases the plants in the same row can also be very close together) spaced enough from each other to allow a small tractor to pass various tools for some necessary workings.

For example, treatments against the onset of particular diseases can be performed by spraying appropriate substances dispensed with diffusers, typically sprayers, pulled by a tractor.

Other more complex operations, such as for example the harvest, can be carried out with very particular machinery, called straddle machines, which induce a strong vibration to the plants, in order to detach the berries which are collected with special trays. In this case, however, the operation is optimal on plants that have fairly homogeneous and predetermined dimensions. The just mentioned examples concern cases of automation in which it is not necessary, for the machinery, to have sensors that detect the shape and position of the plants, or the position of the fruits on the individual plants. In fact, the knowledge that the crops are arranged in a standard mode is sufficient for these machines. Flowever, the availability of more accurate information could bring numerous benefits. For example, when an operator sprays the substances for a treatment against some pathogens by hand, the treatment is carried out only where it is actually needed, and not indiscriminately towards the side where the plants are located, far fewer substances are used and, above all, the quantity of substance that is dispersed in the ground is significantly reduced. It is noted that these substances, with which fruit plants or vineyards must be treated, typically, are chemical products necessary to protect crops from diseases that would compromise them, but they are often substances with potentially harmful effects too, and therefore it is very important to limit their use, only to what is strictly necessary. To date, there are very few, and not very widespread, machines that make use of sensors to perform their work; and when they are used, to acquire information regarding the plants on which these processes must be carried out, some technologies are generally used that can be traced back to three families: video analysis technologies, laser sensors, and infrared sensors. The sensors based on video analysis allow to obtain quite complete information, in theory similar to the information available to a farmer when he does his work by hand. They are sensors that can potentially support robotic machinery with manipulation tools, or tools suitable for performing even very localized processes. For example, CN108633482 (A) - “Fruit picking flight vehicle” - Du Juan et al. 2018/10/12 - teaches to pick fruit with a flying vehicle (i.e. , a robotic drone), possibly also from wild plants. CN108633482 (A) is cited as an emblem of a use of video technologies to support fruit harvesting, but obviously there are also simpler applications; what is important to underline is the great theoretical potential of sensors based on the video analysis, which, on the other hand, also have numerous contraindications.

In fact, these are quite complex and therefore expensive technologies, which require heavy processing and consequently require adequate computing power. In fact, an acquired visible image must first be subjected to a recognition program of the acquired scene, and only after that it is possible to deduce the information required by the machines that perform the job.

Furthermore, outdoor image shooting is very sensitive to light conditions, the optical system to be used (lenses, focusing mechanics and related control electronics) must therefore be of high quality, and the lenses must be maintained clean, even in an environment of use which is particularly predisposed to dirty the equipment.

The family of technologies based on laser sensors, on the other hand, certainly offers simpler and cheaper solutions, but they produce poorer information as they basically operate like a radar, and are essentially able to detect and locate the presence of indistinct masses. CN109764809 (A) - “Method for measuring fruit tree canopy volume in real time on the basis of two-dimensional laser sensor” - Liu Qiang et al. 2019/05/17 - teaches a method to detect with good precision the space occupied by the crown of a fruit tree in order to automatically orient the spray of treatment substances. The teachings of CN109764809 (A), solve a first technical problem with a relatively mature and fairly cheap technology, however laser technologies also have limitations: first of all, that of not distinguishing the various parts of a plant, for example the foliage from the branches and fruits. In particular, the use of laser technologies does not allow to effectively detect the health of a plant, adequately evaluating how thick the foliage is and how many fruits are present. Even with regard to applications based on laser technologies, the teachings of CN109764809 (A) are cited as emblematic of a possible exploitation of the technology in the same context of application of the present invention. In fact, there are some applications, even already on the market, which exploit the technology of the so-called laser sensors to detect the presence of a plant, as well as its shape. The last family of technologies used to acquire information relating to plants consists in the use of infrared sensors. It should be noted that the infrared spectrum is exploited to create a large variety of sensors, including sensors that only have the function of receiving infrared radiation (passive sensors), such as thermal sensors or motion sensors (also the latter ultimately rely on the detection of temperature differences). In the case of applications in agriculture, for the acquisition of information relating to individual plants, however, active sensors are particularly useful, whose operating principle is once again similar to the operation of radar, and therefore it is required the emission of a radiation, of which, the reflected component is then detected. Sensors of this type are beginning to be exploited in agricultural applications as they are able to detect and distinguish the presence of luxuriant vegetation. A rather interesting form of application is that proposed by the Californian company Trimble Inc., which produces numerous solutions to support extensive mechanized agriculture. The most efficient application that makes use of active infrared sensors involves mounting these sensors on long arms that are passed horizontally over relatively low plants. A sequence of sensors mounted along the entire arm detects information on the vegetative state of the underlying plants and, through dispensers mounted on the same arms, treatments, nutrients and water can be sprayed in calibrated quantities according to the needs of the underlying plants. This application makes use of infrared radiation that does not require particular directivity, the sensor is passed very close to the plants, and the only information that is processed is that relating to the vegetative state, i.e. , how luxuriant the seedlings are in relation to their phase of expected growth. In fact, it is not important to have spatial information (location of the plant or some of its parts), nor information about its shape.

This information relating to the vegetative state of a plant is certainly very interesting, and can be obtained in a very direct way as a function of the reflection response of the infrared radiation. Other methods based on optical analyzes would be far more complex, in addition to the fact that they would be strongly conditioned by the light conditions and by the other contraindications already mentioned in relation to optical sensors in the visible range.

This technology based on active sensor in infrared radiation, while being very promising, is essentially applied only to support ground crops in which the plants are structurally quite simple and their spatial conformation is known (for what is needed) from other ways. On the other hand, it is not significantly applied to support the treatments to be carried out on other types of crops characterized by more complex plants, such as orchards and vineyards. This is mainly because it has not yet been possible to develop a technology with suitable characteristics of economy and ease of use, so as to be proposed it in the real agricultural context of orchards and vineyards; or contexts in which it is often very important to acquire information of a certain detail regarding the state of individual plants.

In general, on orchards and vineyards, many jobs continue to be carried out manually, or, when they are mechanized, such as in the case of carrying out the necessary treatments of the plants, the treatment is done with less precision and with waste of product; with the double contraindication of wasting a product that still has a certain cost, and of unnecessarily scattering on the ground products, which often also have a polluting effect that should be as limited as possible.

Purpose and summary of the invention

The general purpose of the present invention, therefore, is to indicate an efficient sensor, and achievable at low cost, which allows to acquire information on the vegetative state of vertically growing plants such as the trees of an orchard or the vines of a vineyard.

In particular, it is important to have information that distinguishes the presence of leaves, flowers, branches and fruits, and that spatially localize these different parts of the plants.

Another purpose of the present invention is to indicate a sensor that can easily be towed even by a small tractor, which can easily pass between the rows of an orchard or a vineyard, without said sensor having to be moved by specific motors or other mechanisms, in order to produce a sort of two-dimensional image of an entire row of plants.

The objects of the present invention can be achieved by making a sensor device having an elongated shape (typically consisting of a bar on which the sensory components are mounted), suitable to be mounted vertically, and rigidly, on a self-moving or towable means of locomotion, comprising a plurality of transceiver elements of infrared radiation (i.e. , active infrared sensors), being said sensor device characterized in that said transceiver elements: are arranged in an aligned manner, so that when said elongated sensor device is mounted vertically, said transceivers are also vertically aligned, transmit in narrow band on at least two distinct frequencies of the infrared spectrum, transmit directionally with radiation lobes, whose 3 dB width, at a distance of one meter from the emission point, does not exceed 10 cm.

The main advantage of the present invention consists in the fact that, as will be better explained below, a sensor device made according to the teachings of the present invention satisfies all the main requirements for which it was conceived.

Brief Description of the Drawings

This invention also has further advantages, which will become more evident from the following description, from some examples of practical embodiments, which illustrate further details, from the attached claims, which form an integral part of the present description, and from the attached figures in which: Figure 1 schematically shows a sensor device according to the invention in a typical application context. Figure 2 schematically shows a graph that highlights how the data collected by the sensor can be directly interpreted to recognize the various parts of a plant.

Detailed description

Figure 1 schematically shows a context of use of the sensor device according to the invention, depicted within a dashed line, and indicated as a whole with the number 100.

Said sensor device 100 is essentially composed of: calculation means, suitable to also support data storage functions and numerical control functions, and indicated in Figure 1 with the number 130, an elongated support structure, indicated in Figure 1 with the number 120, and a plurality of infrared emission active sensors, indicated in Figure 1 with the number 110 (the number in Figure 1 indicates only the active sensor positioned in the top of the row, but all the others are substantially the same). Said calculation means 130 perform various functions, among which the control of the sensors 110, the storage of the acquired data, as well as their processing.

Furthermore, said calculation means 130 are also suitable for executing any further programs functional to the control of any other systems designed to operate in synergy with the sensor device 100 according to the invention, by carrying out specific jobs on the plants of an orchard or a vineyard, making use of the information acquired by the sensor device 100 according to the invention.

Said support structure 120 is designed to be mounted vertically on a generic trolley or base, and has the function of rigidly supporting a plurality of active infrared sensors, keeping them substantially aligned so as to form a vertical line.

It should be remembered that with infrared emission active sensor we mean a sensor that includes a transmitting component of a radiation in the infrared spectrum and a receiving component suitable for detecting the reflected component of the transmitted radiation (when this is reflected by the surface towards which it was been directed).

Said infrared emission active sensors 110 represent an essential distinctive element of the present invention. In fact, compared to sensors of the same technology applied in agriculture, the active sensors 110 have some technical characteristics that differentiate them from the sensors currently used, and make them effective with respect to the purposes of the invention.

Furthermore, the simplicity with which said active sensors 100 can be made, as well as functional effectiveness, is an essential technical requirement to achieve practical success in the use of the invention. Specifically, the functional characteristics required for the implementation of the present invention are well combined in order to be able to technically realize these optical sensors 110 in a fairly simple way, so as to obtain production costs compatible with the market to which they are addressed.

The first technical feature consists in the fact that the infrared emission is a narrow band emission, on at least two central frequencies; even if, in the preferred implementation forms, already developed today for numerous types of plants, three radiations on three different frequencies are transmitted, and in some cases four radiations on four frequencies.

As it will be clarified later on, the narrow band emission is essential to distinguish the type of plant tissue irradiated by the infrared emission, however it represents a technical characteristic that allows to emit a very directive beam without resorting to expensive focusing optics (which would limit the concrete diffusion of the invention in practical use).

In a form of implementation tested with considerable success in an apple orchard (therefore with fruits of the size of an apple) there were used sensors with a beam width (radiated by each sensor 110) of about 7 cm at a distance of 2 meters from the point of emission. Such sensors 110, with these directivity performances, have been obtained very easily (and therefore at absolutely acceptable costs) by making them emit on three frequencies in the infrared spectrum: with wavelengths at 660 nm, 780 nm and 890 nm (some considerations on the advisability of using these frequencies will be discussed later).

The geometric characteristics of the just described sensor device 100 allow to conclude how, taking care to arrange in an aligned and vertical way a row of active infrared emission sensors 110, it is possible to illuminate a plant (or any other body), with a beam of substantially linear and vertical footprint. Furthermore, by horizontally dragging said sensor device, for example because it is mounted on a trolley, towed or self-moving, it is possible to illuminate, over time, all the points of a row of plants that are visible laterally, operating a sort of lateral "scanning" of this row of plants.

It should be noted that the precise verticality of the alignment of the sensors 110 is not essential in itself, in theory the sensors could also be aligned obliquely (for example by mounting the support structure 120 obliquely); what matters is that each sensor is at a different height, so that the sensor device 100 as a whole illuminates the row, at all heights, and, as a result of its horizontal dragging, ultimately illuminates the entire row.

However, for reasons of simplicity, it is clear that, in the preferred embodiment, the sensor is mounted in a vertical position.

Obviously, all the data acquired by the sensor device 100 must be collected and stored, to reconstruct a two-dimensional map, or a sort of image (understood in the sense that will be clarified later), of the side view of a row of plants. The two- dimensionality of this map being given by the multiplicity of sensors 110 for the vertical dimension, and by the multiplicity of different detections over the time for the horizontal dimension.

The definition of the resulting image (or of the map, given that, as explained, it is not a real image in the visible field, but a representation of another type) depends on the directivity of the individual sensors 110 and on the detection frequency of each moving sensor 110, in relation to the horizontal drag speed.

Even if the information on the dragging speed can be acquired directly from the means that carries the sensor device 100, in many embodiments it is very useful to integrate a speed sensor in the sensor device 100 as well, since the frequency of acquisition, by the elementary sensors 110, must be calculated depending on this speed of dragging.

Figure 1 shows a typical application case, in which the sensor device 100 is mounted on a trolley, indicated with the number 302, pulled by a tractor, indicated with the number 301. During the movement of the sensor device 100, it acquires a two- dimensional image of everything on the side of the path illuminated by the sensor device 100, in a height range corresponding to the length of the sensor device 100 itself.

Finally, in Figure 1 , the number 200 indicates a system for carrying out a generic treatment to be performed on the plants: it is clear that this system 200 can make use of the two-dimensional image acquired from the sensor device 100, to carry out the envisaged treatment with greater accuracy.

It is observed that the control of this system 200 for carrying out a generic treatment (which in many cases could be exercised directly by the same calculation means 130 which also control the sensor device 100), also needs to know in time real is the dragging speed with which said sensor device (100) moves horizontally. This speed data, therefore, represents information of considerable importance.

Given the structural simplicity of the sensor device according to the teachings of the present invention, it is obviously also possible to set up a tractor so that it horizontally drags two sensor devices 100, oriented towards both sides of the path. In this way, with a single passage between two rows of plants, it is possible to

Other implementation variants obviously concern the definition of the image to be acquired. The example of implementation form shown in the piece of description above allows a very good definition, less than a decimeter with detections at a distance of 2 meters. Obviously, higher definitions are certainly improvements, in general, but the invention can also be advantageously used also by resorting to simpler systems, with smaller definitions (and therefore less expensive to obtain). In general, it can be said that active sensors capable of transmitting with sufficiently directive radiation lobes, such that the 3 dB width of the lobe at a distance of one meter from the point of emission is not more than 10 cm, can be considered usable systems for making acceptable embodiments of the present invention.

The number of frequencies on which to transmit, as well as the specific frequencies used, is also the subject of possible variants of the same invention, even if the solution with three frequencies has proved to be suitable for satisfying the purposes of the invention very well. From a strictly technological point of view, various implementation choices can also be reported regarding the hardware structure of the individual active sensors, to ensure narrowband transmission on two or more carriers. A simple choice consists in making single-carrier transmitters, and given the small size of these devices, they can be arranged side by side, close to each other, so that when the sensor device is mounted vertically, the individual emitters are arranged horizontally, so that they are all at the same height, and oriented so as to illuminate the same area of the plant. A different implementation choice involves directly making multi-carrier emitters, but, obviously, one or the other choice does not modify the inventive principles underlying the present invention.

The way in which the aforementioned multi-carrier lighting is exploited in order of achieving the purposes of the present invention is now better explained with the aid of Figure 2.

In fact, the invention is based on the exploitation of a discovery: namely the discovery that the way in which the individual infrared frequencies are reflected by different tissues of vegetal nature is different. Obviously, the reflection of infrared radiation depends on many factors, some even contingent, including the temperature and orientation of the reflecting surface, etc. however, it was noted that the structure of the fabric also introduces an effect that differentiates the reflections, more accentuated on some wavelengths than others. This discovery was exploited to detect a sort of “images” with very simple sensors, such as infrared sensors. Given the experimental nature of the aforementioned discovery, the development of the analysis procedure required the execution of a large number of tests in many environmental conditions.

The result that has been achieved has shown that, by illuminating four different types of plant tissues, with three different frequencies in the infrared spectrum, and reporting in a three-dimensional coordinate system the intensity of the reflected radiation of each radiation at different frequency, the points expressed on this three- dimensional coordinate system accumulate in different zones, depending on the type of fabric from which the radiation is reflected. Four types of fabrics have been observed and distinguished: green foliage, the woody texture of the branches, flowers and fruits.

Figure 2 shows a graph in which some points corresponding to some experimental acquisitions are positioned. It is noted that some points accumulate mainly on the "f1 , f2" plane (evidently the tissue affected by this radiation does not reflect the frequency "f3" well); it is also noted that almost all these points are marked with the same symbol (in the case of Figure 2, with the symbol “x”), meaning that they are signals reflected by the same tissue. In the same way, some points tend to accumulate near the plane "f1, f3", and also in this case these points close to the plane "f1 , f3" are marked with the same symbol (in the case of Figure 2, with the symbol "o”), and also in this case they are points associated with reflections all coming from the same fabric. Finally, there are other points that tend to accumulate near the plane "f2, f3", and also in this case they are marked with the same symbol, different from the other symbols (in the case of Figure 2, with the symbol "+"), because they refer to reflections coming from yet another fabric. The example of Figure 2 clearly shows how each reflection is different, as there are few overlapping points (or in any case very close), however, by appropriately selecting three different frequencies, in the infrared, with which to illuminate the individual tissues, it is possible to obtain reflections that are interpretable as coming from different plant tissues. Therefore, the receiving part of the active sensors 110 is able to return a value that represents a pixel of the two-dimensional image constructed as explained with the aid of Figure 1.

In an essential implementation form, the value assumed by each pixel belongs to a set of at least five values corresponding to the prevailing presence in that pixel of a branch, foliage, fruit or flower, in which the fifth value corresponds to the absence of reflection, due to the fact that the radiation does not affect any part of the plant.

In other embodiments these values can also be in greater number so as to also offer other information on the health of the plant. In other words, the sensor device 100 can reconstruct ever richer images (from an information point of view) concerning the plant that is "scanned".

It is reiterated that the present invention is based on the application of a discovery, which certainly has theoretical justifications, but which, in its detail, has been experimentally analyzed, therefore the sensor device 100 is certainly subject to further refinements aimed at always defining greater precision in the illuminating frequencies to be used and in the threshold areas that allow to discriminate between one tissue and another, or to interpret the acquired values also to derive information about the general state of the plant, without limiting itself to recording the mere presence of fruit, leaves, flowers and branches.

Variants and Concluding Remarks Ultimately, the sensor device 100 according to the teachings of the invention can be used to acquire a two-dimensional image, or map, of a row of plants, without going through the acquisition of an image in the visible field which would involve the use of optical sensors, which presents various problems; instead reconstructing equivalent images (i.e. maps describing a row of plants seen from its side) which are very significant from the point of view of their information content, as just explained above.

In fact, these images, acquired in this way, do not need to be reprocessed through recognition programs, but directly contain the information that is necessary for many types of applications.

The known technique, in general, has tended to solve the problem of the automation of agricultural processes through the imposition of increasingly standardized cultivation systems; however, although we try to control the position, size and shape of the plants, in order to facilitate their mechanized processing, a significant number of processes to be carried out on vineyards and orchards require, in order to achieve real optimizations, to know with greater precision the conformation and state of each individual plants, also with regard to all those details that make them different from each other.

The use of active infrared radiation sensors in the agricultural sector, despite being a recently introduced technology, is not an absolute novelty to detect information on the vegetative state of some plants, but this technology had never been used to detect real images (different, of course, from simple thermal images that could be detected with passive sensor arrays).

The invention is therefore based on a new use of the technology of infrared radiation active sensors, aimed at solving various technical problems which, if addressed with known solutions, actually have not yet satisfactory performances. Such new use requires the conception and implementation of a new type of sensor device, built with specific new features, such as the use of a very directive radiation, in narrow band, and on a plurality of carriers. Furthermore, the geometric structure of the sensor device must also have particular characteristics that make it suitable for performing a sort of "scanning" of entire rows of plants; in a simple way, and without the need to resort to complex handling mechanisms. The sensor device 100 according to the invention, as mentioned, is of particular interest as it allows to optimally automate some agricultural processes on orchards and vineyards, which today are performed manually or with significant inefficiencies.

Therefore, it is an invention that naturally lends itself to being conceived as an element that must in turn be integrated into more complex systems aimed at automating specific agricultural processes.

Consequently, the invention lends itself to incorporating variants or improvements related to the specific processing performed by the individual systems in which it can be integrated. For example, for some processes it could be very important to know with good precision also the distance of the plant from the sensor, at the time of detection. It is noted that an approximate figure is certainly already deducible from the power with which the reflected radiation is received, however the approximations that can be obtained are of the order of decimeters, and may not be adequate for certain applications. The technology makes available rangefinders at very low cost, and of a size that can be easily integrated into sensor devices 100 according to the invention.

Therefore, in a form of implementation of a certain utility (in particular for certain applications) the sensor device 100 according to the invention can also integrate one or more telemeters, also possibly controlled by the calculation means 130, and suitable for acquiring in real time the distance of the surface reflecting the infrared radiation.

Possible further variants may then be linked to the evolution of the other components of these complex systems on which the sensor device 100 can be integrated; so that each element of each subsystem of said complex systems (one of which is precisely the sensor device 100 according to the invention) can be subject to variations that allow it to be used for a plurality of functions, which at the moment are not yet definable, or for which the need is not yet felt.

Therefore, especially in the context of these evolutionary scenarios, the invention seems to lend itself to incorporating and supporting further development and improvement efforts, capable of improving the performance of the systems in which it can be integrated.

Many further developments can therefore be made by the man skilled in the art without thereby departing from the scope of the invention as it results from this description and the attached claims, which form an integral part of this description; or, if said developments are not included in the present description, they may be the subject matter of further patent applications associated with the present invention, or dependent on it.

Claims

1. A sensor device (100), having an elongated shape, suitable to be mounted vertically, and rigidly, on a self-moving ortowable means of locomotion (302), comprising a plurality of transceiver elements (110) of infrared radiation, arranged in an aligned manner, so that when said elongated sensor device (100) is mounted vertically, said transceivers (110) are also vertically aligned, and being said sensor device (100) characterized in that: a. the transmission performed by said transceiver elements (110) is in narrow band, and on at least two distinct central frequencies of the infrared spectrum, b. the transmission performed by said transceiver elements (110) is directional, and has a radiation lobe, whose 3 dB width, at a distance of one meter from the emission point, does not exceed 10 cm; and c. said sensor device (100) is configured to acquire a sequence of vertical images of longitudinal shape, each consisting of a number of pixels corresponding to the number of said transceiver elements (110), and said pixels assume at least five different values depending on whether they represent an area in which the plant tissue they reproduce is mainly made up of: i. leaves, ii. a flower, iii. a fruit, iv. a woody part, or v. absence of any plant tissue.

2. Sensor device (100) according to claim 1 , wherein said transceiver elements (110) transmit in narrow band on three distinct frequencies of the infrared spectrum.

3. Sensor device (100) according to claim 1 , wherein said transceiver elements (110) transmit in narrow band on four distinct frequencies of the infrared spectrum.

4. Sensor device (100) according to claim 1, which also includes calculation means (130), suitable for supporting also data storage functions and numerical control functions.

5. Sensor device (100) according to claim 4, which also includes a speed sensor, suitable for acquiring in real time the dragging speed with which said sensor device (100) moves horizontally.

6. Sensor device (100) according to claim 5, wherein the data acquisition frequency is calculated by means of said calculation means (130), as a function of the dragging speed with which said sensor device (100) moves horizontally.

7. Sensor device (100) according to claim 6, in which said calculation means

(130) are arranged to store the acquired data by composing them in the form of a map representative of a row of plants seen from its side, wherein each point of this map corresponds to a piece of information that indicates the presence in that point of a part of a plant, distinguishing whether this part corresponds to a wooden fabric, to foliage, to a fruit or to a flower.

8. Sensor device (100) according to claim 4, which also includes a distance sensor, or a rangefinder, suitable for acquiring in real time the distance at which the surface reflecting the infrared radiation is located.