Optical sorting of non-cerebaric fruits by combination of electromagnetic radiation

Description

[0001] The invention relates to a non-destructive method for determining the maturity and quality of a non-cereal fruit, a method for sorting non-cereal fruits and an apparatus for sorting such non-cereal fruits implementing these methods.

[0002] The industrialization of high-throughput sorting techniques for agri-food products from crops is essential for mass production, particularly for sorting products according to their maturity.

[0003] It is known that product maturity can be assessed by determining the chromophore compounds contained in plant cells. The maturity in question can primarily refer to phenolic maturity. This type of system offers the possibility of being used online and integrated into a production line.

[0004] It is known for example from JPH11621A and KR20030088706, processes for sorting rice grains by exposure to electromagnetic radiation and signal analysis.

[0005] Similarly, in EP0898701 a process for determining the maturity and quality of seeds by electromagnetic irradiation is reported.

[0006] Thus, these processes are particularly well suited to fruits / cereal products, which typically have a limited number of maturity markers.

[0007] One of the advantages of these processes is the speed of the rice seed or grain selection stage, making them particularly attractive for implementation in high-volume production lines. Furthermore, chlorophyll fluorescence is a determining factor in the processes described.

[0008] Regarding non-cereal fruits, G. Agati et al. ("Nondestructive evaluation of anthocyanins in olive (Olea europaea) fruits by in situ chlorophyll fluorescence spectroscopy"; J Agric Food Chem. 2005, (5), 1354-63) reported that anthocyanins in olive (Olea europaea L.) fruits at different levels of pigmentation were evaluated non-destructively by measuring chlorophyll fluorescence. This was achieved using a method based on comparing the chlorophyll excitation spectra of different olives with varying pigmentation, ranging from green to green-red, red-violet, and violet. However, this technique is limited to fruits at low or medium maturity. Irradiation with green and red light does not allow for the assessment of the full range of maturity.

[0009] However, WO 00 / 02036 describes the use of fluorescence to determine the presence or chlorophyll content of a plant, particularly berries. However, for the ripest berries, the chlorophyll level can be very low. This weak chlorophyll fluorescence signal may even be insufficient for detection, potentially leading to sorting errors and a decrease in the quality of the sorted product. The technique therefore needs improvement. Furthermore, in WO 00 / 02036, the described power levels are insufficient, especially in the green range: the described technology thus appears limited to screening a so-called mature population.

[0010] Document EP3379246A1 discloses a non-destructive method for determining the maturity and quality of a non-cereal fruit, which may include two steps of exposing said fruit to at least two different electromagnetic radiations and measuring their return signals. However, document EP3379246A1 has practical limitations in implementation, particularly regarding the quality of the signals obtained, which has a direct impact on the quality of the isolated fruits.

[0011] The aim of the invention is therefore to overcome the disadvantages of the prior art and thus aims in particular to overcome the difficulties of sorting non-cereal fruits on an industrial scale.

[0012] Furthermore, in general, the fluorescence signal from chlorophyll is dependent on the excitation power and the chlorophyll level. Low chlorophyll levels require lighting powers that present a technical (i.e., a technological) and economic challenge. SUMMARY OF THE INVENTION

[0013] To this end, a first aspect of the invention relates, in its broadest sense, to a non-destructive method for determining the maturity and quality of a non-cereal fruit according to claim 1 and comprising at least two steps of exposing said fruit to at least two different electromagnetic radiations and measuring their return signals, in which: one of the exposure steps of said fruit includes the provision of at least one wavelength capable of causing a chlorophyll fluorescence signal and a measurement, where appropriate, of said chlorophyll fluorescence signal; and one of the exposure steps of said fruit to at least one electromagnetic radiation includes the provision of at least one selected wavelength capable of generating a reflection signal on said fruit, and a measurement, where appropriate, of said reflection signal, characterized, in particular, in that said at least one wavelength capable of causing chlorophyll fluorescence is between 590 nm and 660 nm.

[0014] Indeed, other maturity markers besides chlorophyll are known (such as anthocyanins), which contribute to producing the characteristic colors of fruits. A mutual contribution of the different maturity markers thus occurs, resulting in variations in the levels of these different markers in the fruit. The present invention uses this concept to aid the maturity assessment process of non-cereal fruits, and thus help to increase the accuracy of industrial sorting of such fruits.

[0015] One of the principles of the present invention is to correlate a signal from an electromagnetic reflection chosen according to the fruit being sorted with a chlorophyll fluorescence signal. The signals obtained are thus complementary in that they provide different information useful for sorting. Therefore, thanks to the process according to the present invention, selection becomes more precise and sorting of higher quality.

[0016] In particular, the wavelength ranges likely to induce chlorophyll fluorescence are between 590 nm and 660 nm, allowing for greater specificity of irradiation and therefore greater certainty of the results. One of the advantages of the invention is the ability to distinguish two different but complementary signals with respect to the characteristics of a particular fruit, specifically grapes. Indeed, it has been discovered, particularly for grapes, that anthocyanins do not mask chlorophyll fluorescence at the wavelength ranges emitted when chlorophyll fluoresces upon excitation at wavelengths of 590 nm and 660 nm.Thus, when the excitation wavelength is between 590 and 660 nm, the emitted signal (chlorophyll fluorescence) is complementary to the reflection signals of the anthocyanins present and the whole of the data collected makes it possible to better characterize the fruit qualitatively with regard to the known state of the art techniques.

[0017] The object of the present invention also relates to a non-cereal fruit sorting apparatus according to claim 8, the apparatus comprising: a non-cereal fruit feeding device, optionally at least one means of removing objects that are not fruits; at least one radiation system configured to generate at least one electromagnetic radiation; at least one detection zone configured to measure and optionally analyze the return signals from said at least one radiation system; a fruit separation system activatable according to the return signals, characterized in that: said at least one radiation system is configured so as to expose said fruits to at least two different electromagnetic radiations; one of said at least two electromagnetic radiations includes at least one wavelength capable of causing chlorophyll fluorescence; one of said at least two electromagnetic radiations includes at least one wavelength chosen to generate a reflection on said fruit, such as 750 nm;said at least one radiation system and the detection zone are arranged so that: at least two return signals generated by said at least one radiation system are captured on the same detection zone, or are on the same side of the fruit (otherwise, it would be possible that the fruit support is transparent for example); characterized in particular, in that said at least one wavelength capable of causing chlorophyll fluorescence is between 590 nm and 660 nm, and in that the measurements of the chlorophyll fluorescence signal and the reflection signal are carried out by the same sensor.

[0018] Thus, the apparatus according to the present invention improves the sorting of non-cereal fruits by using a signal complementary to the signal obtained by the fluorescent excitation of chlorophyll. This complementary signal is specifically calibrated to the type of fruit to be sorted. Furthermore, unlike apparatus used to sort certain cereals such as rice, the radiation systems and detection zones are located on the same side of the fruit. Indeed, rice is partially translucent, which necessitates a different configuration of the detection and emission systems compared to other non-cereal fruits.

[0019] The present invention also makes it possible to identify objects that are not fruits: if no chlorophyll fluorescence signal is detected and the reflection signal is not characteristic of the fruit being analyzed, then it is a foreign body that can be removed. Other methods of distinguishing objects other than fruits can be used (weight-based measurement systems, for example).

[0020] The object of the present invention further relates to a method for sorting non-cereal fruits according to claim 9, the method comprising the following steps: individually bring each non-cereal fruit to at least one exposure zone, expose said fruit in said at least one exposure zone to at least two different electromagnetic radiations, receive and analyze the return signals from the exposure stage, separate said fruits into different classes according to their individual signals, in which: one of the at least two different electromagnetic radiations includes at least one wavelength capable of causing chlorophyll fluorescence as a return signal; one of the at least two different electromagnetic radiations includes at least one wavelength chosen to generate reflection on said fruit as a return signal; and the values ​​that define the classes are pre-established or chosen based on a distribution of chlorophyll reflection and fluorescence signals emitted by a sample of non-cereal fruits possessing known properties, characterized in particular, in that said at least one wavelength capable of causing chlorophyll fluorescence is between 590 nm and 660 nm, and in that the measurements of the chlorophyll fluorescence signal and the reflection signal are carried out by the same sensor.

[0021] Thus, this process allows the use, if desired, of a predefined threshold value, or the selection of this sorting threshold for non-cereal fruits based on a preliminary study of a sample of the harvest to be sorted. The present invention therefore makes it possible to maintain, if desired, a constant, or "standardized," level of quality for the sorted fruits, or conversely, to adapt the sorting to the reality of the harvest and thus adjust, for example, the ratio between sufficient production yield and the quality of the harvested fruits.

[0022] The invention (sorting methods and apparatus) is therefore particularly well suited for sorting fruits such as grapes in the context of the production of non-cereal fruit products such as drinks like wine, fruit juices, coulis, etc., dairy products including fruit or parts of fruit, such as fruit yogurts, or fruit desserts such as pies, cakes, mousses, etc., which are particularly sensitive to the quality of the fruit used. DEFINITIONS

[0023] For the purposes of this invention, "non-cereal fruit" means a fruit that is not a cereal fruit. For the purposes of this invention, "cereal fruit" means the botanical definition of a fruit, that is, a grain, or seed, of a cereal. Cereals are monocotyledonous plants of the grass family, wild or cultivated, that produce edible grains, primarily used for human consumption, often ground into refined or partially wholemeal flour. Examples of cereals include maize, wheat, rice, barley, sorghum, oats, rye, millet, and triticale.

[0024] For the purposes of this invention, a "detection zone" is defined as a surface sensitive to at least one electromagnetic radiation (i.e., light) and configured to convert a light signal into a usable signal, such as an electrical signal. Typically, such a detection zone is a camera, such as a digital camera like a CCD or CMOS camera. The camera may be matrix or linear. The detection zone may further be equipped with one or more light filters to eliminate light signals that are irrelevant to the present invention and thus improve the signal-to-noise ratio.

[0025] In the context of this invention, the term "return signal" refers to the signal from the fruit to a sensor. This return signal may be a chlorophyll fluorescence signal or a reflection signal.

[0026] In the context of the present invention, "fluorescence signal" means the signal resulting from the excitation of chlorophyll.

[0027] In the context of the present invention, "reflection signal" means the light signal which is not absorbed by the non-cereal fruit and is therefore reflected.

[0028] By "same side of a fruit", it is understood in the context of the present invention to mean a face of the fruit seen from a single point outside the fruit. DETAILED DESCRIPTION • Method for determining the maturity and quality of a fruit:

[0029] The present invention relates in particular to a non-destructive method for determining the maturity and quality of a non-cereal fruit, such as grapes, comprising: a first step of exposing said fruit to at least one electromagnetic radiation and measuring the return signal, a second step of exposing said fruit to at least one electromagnetic radiation different from that of the first step and measuring the return signal, in which: One of the exposure steps of said fruit includes the provision of at least one wavelength capable of causing a chlorophyll fluorescence signal and a measurement, if applicable, of said chlorophyll fluorescence signal; and one of the exposure steps of said fruit to at least one electromagnetic radiation includes the provision of at least one selected wavelength capable of generating a reflection signal on said fruit, and a measurement, if applicable, of said reflection signal.

[0030] According to the invention, the method according to the present invention is characterized in that the measurements of the chlorophyll fluorescence signal and the reflection signal are carried out by the same sensor.

[0031] There are several advantages to using a single sensor to collect both return signals according to the invention. For example, a single sensor reduces the cost of the equipment. A single sensor also reduces maintenance requirements. A single sensor allows for a more compact design. Furthermore, a single sensor ensures that the chlorophyll fluorescence and reflection signals originate from the same side of the analyzed fruit, leading to a more accurate correlation between the two signals and thus a better assessment of the fruit's overall quality.

[0032] Thus, advantageously, the method according to the present invention can be characterized in that the sensor is a camera optionally provided with a bandpass filter, for example a bandpass filter calibrated at 750 nm.

[0033] Thus, in a particular embodiment, the method according to the present invention can be characterized in that, on the one hand, the supply of at least one wavelength capable of causing a chlorophyll fluorescence signal, and on the other hand the supply of at least one selected wavelength capable of generating a reflection signal on said fruit, are sequenced (in time separately) and / or synchronized, with optionally at least one pause time to allow measurement of background noise.

[0034] In a particular embodiment, the radiation systems include one or more light-emitting diode (LED)-based sources.

[0035] Preferably, the method according to the present invention can be characterized in that said at least one wavelength chosen to generate a reflection on said fruit is chosen with a wavelength significantly greater than the excitation wavelength of chlorophyll.

[0036] By "significantly greater than the excitation wavelength of chlorophyll", it is understood in the context of the present invention to mean a wavelength which does not generate chlorophyll excitation capable of being detected (in particular in signal intensity) by the sensor used.

[0037] More preferably, the process according to the present invention can be characterized in that the wavelength significantly greater than the excitation wavelength of chlorophyll is between 700 nm and 900 nm, preferably 750 nm plus or minus 10 nm.

[0038] For example, the process according to the present invention can be characterized in that the wavelength significantly greater than the excitation wavelength of chlorophyll is between 710 nm and 850 nm, between 720 nm and 800 nm, between 730 nm and 780 nm or between 735 nm and 765 nm.

[0039] More preferably, the process according to the present invention can be characterized in that said at least one wavelength capable of causing chlorophyll fluorescence is between 550 nm and 680 nm, preferably between 590 nm and 660 nm, more preferably between 625 nm plus or minus 10 nm.

[0040] For example, the method according to the present invention can be characterized in that said at least one wavelength capable of causing chlorophyll fluorescence is between 590 nm and 660 nm, between 600 nm and 650 nm, or between 610 nm and 640 nm, in particular 625 nm plus or minus 10 nm.

[0041] In a particular embodiment, the process according to the present invention can be characterized in that the wavelength significantly greater than the excitation wavelength of chlorophyll is between 700 nm and 900 nm, preferably 750 nm plus or minus 10 nm, and in that said at least one wavelength capable of causing fluorescence of chlorophyll is between 550 nm and 680 nm, preferably 625 nm plus or minus 10 nm.

[0042] The advantage is being able to distinguish two different but complementary signals regarding the characteristics of a particular fruit.

[0043] For example, a fruit like a grape contains at least two markers of ripeness: chlorophyll and anthocyanins. Chlorophyll can be excited with an electromagnetic wave with a wavelength between 550 and 680 nm, and the fluorescence thus generated can be between 720 and 770 nm, such as 750 nm plus or minus 10 nm. When the excitation wavelength is between 590 and 660 nm, the emitted signal (chlorophyll fluorescence) is complementary to those of the anthocyanins present and allows for the qualitative characterization of the fruit.

[0044] Thus, one of the distinctive features of the present invention is that chlorophyll is detected by fluorescence and the reflection signal provides additional information (e.g., presence of an object and shape of the irradiated object).

[0045] To measure the ripeness of a fruit like a grape, the fruit can undergo a first sequential red excitation (e.g., 625 nm) and the emission of a second light beam, e.g., 750 nm. The first return signal is therefore characteristic of chlorophyll fluorescence (e.g., around 750 nm) and the second return signal resulting from the reflection (e.g., around 750 nm) of the second emission at 750 nm on the fruit, characteristic of the presence of anthocyanins (second marker of ripeness).

[0046] In a particular and advantageous embodiment, the support used to bring the fruit into the analysis zone (where it will be irradiated) interacts little or not at all with the electromagnetic waves at the chosen wavelengths. Thus, the signals collected do not originate from the support itself. For example, the support may include a transparent, glazed section to avoid interfering with the signal.

[0047] Thus, the irradiation power (such as in the infrared range) is specifically adjusted to prevent the substrate from saturating the camera, depending on the fruit being analyzed. The return signal from the fruit is not lost in a mass of signal originating from the substrate.

[0048] Furthermore, in the context of the present invention, the residual chlorophyll content of a fruit such as grapes is measured by avoiding the masking effect of anthocyanins after excitation with an appropriate wavelength (e.g. 750 nm).

[0049] In this particular case where the maturity markers are chlorophyll and anthocyanins, it may be particularly advantageous to use a single detection zone (e.g., a single camera) since the wavelengths of the return signals are close, or even identical (e.g., approximately 750 nm).

[0050] In the case of using a single detection zone, such as a camera, it is necessary to carry out sequential excitations, for example by using a bandpass filtered camera, to allow the fluorescence to be measured at a wavelength in synchronization with each excitation, and the return signal from the reflection to be measured at a wavelength in synchronization with each excitation.

[0051] Depending on the process, it may be advantageous for the excitation and emission waves to be high-pass filtered (frequency) to remove wavelengths close to the fluorescence wavelength.

[0052] Depending on the method, it may be advantageous for the system to continuously calibrate itself by including a measurement sequence without excitation to discriminate against background noise.

[0053] Thus, the method according to the present invention can be applied to any type of non-cereal fruit containing chlorophyll and a second maturity marker whose return signal obtained by reflection or the return signal obtained by excitation differs from the excitation and fluorescence wavelengths of chlorophyll. This therefore depends directly on the nature of the non-cereal fruit.

[0054] Preferably, the process according to the present invention can be characterized in that the non-cereal fruit comprises chlorophyll and at least one second maturity marker useful for reflection. Indeed, in the absence of chlorophyll, the second marker will reflect the light (electromagnetic waves) to which the fruit is exposed.

[0055] Thus, preferably, the process according to the present invention can be characterized in that the non-cereal fruit contains at least one pigment (or maturity marker) selected from lycopene, beta-carotene, anthocyanin and lutein, preferably anthocyanin and / or lycopene.

[0056] Preferably, the process according to the present invention can be characterized in that the non-cereal fruit is a fleshy fruit.

[0057] A fleshy fruit is a fruit whose mesocarp has the consistency of pulp, such as berries or drupes.

[0058] Preferably, the process according to the present invention can be characterized in that the non-cereal fruit is chosen from a tomato, a grape, a strawberry, a raspberry, a redcurrant, a cherry, a blackberry, a lingonberry, a blackcurrant, an aronia berry, a cranberry, a black elderberry and a blueberry, preferably a grape such as a red grape.

[0059] Preferably, the process according to the present invention can be characterized in that the non-cereal fruit is a red fruit.

[0060] In the context of the present invention, a red fruit is a berry such as a grape, strawberry, raspberry, gooseberry, cherry, blackberry, lingonberry, blackcurrant, aronia berry, cranberry, elderberry or blueberry, preferably a grape such as a red grape.

[0061] Furthermore, it may be advantageous to add to the process according to the present invention a known complementary sorting technique based on the color of the non-cereal fruit. • Sorting process for non-cereal fruits:

[0062] The object according to the present invention thus comprises a method for sorting non-cereal fruits applying the non-destructive method for determining the maturity and quality of a non-cereal fruit as described above, to which the following steps have been added: individually bring each non-cereal fruit to at least one exposure zone, separate said fruits into different classes according to their individual signals.

[0063] This process is put into practice by the sorting device according to the invention generally described above and in more detail below. • Non-cereal fruit sorting machine:

[0064] The non-cereal fruit sorting apparatus according to the present invention includes in particular a non-cereal fruit feeding device.

[0065] Non-cereal fruits, such as grape berries, are moved by means of a conveyor; the speed of the conveyor ejects the berries at its end along a known trajectory.

[0066] Preferably, the non-cereal fruit sorting apparatus according to the present invention can be characterized in that said apparatus is configured to sort fruits as described above in the process according to the invention, for example selected from tomatoes, grapes, strawberries, raspberries, currants, cherries, blackberries, lingonberries, blackcurrants, aronia berries, cranberries, elderberries and blueberries, preferably grapes such as red grapes.

[0067] The non-cereal fruit sorting apparatus according to the present invention includes in particular at least one radiation system configured to generate one or more electromagnetic radiations.

[0068] In a particular embodiment, the non-cereal fruit sorting apparatus according to the present invention comprises two radiation systems configured to emit at least two electromagnetic radiations.

[0069] The non-cereal fruit sorting apparatus according to the present invention further comprises at least one detection zone configured to measure and optionally analyze the return signals from the radiation system(s). One detection zone and two radiation systems configured so that the return signals reach said detection zone can constitute a non-cereal fruit maturity analysis set. The non-cereal fruit sorting apparatus according to the present invention can comprise several non-cereal fruit maturity analysis sets, each arranged to expose said non-cereal fruits at different angles.

[0070] In a particular embodiment, the distance of the detection zone from irradiated fruit is adapted to have both the largest area for capturing return signals, without the detection means suffering from hazards related to its proximity to the fruit (spraying of juice, dirt or other).

[0071] In addition, each analysis set may include a device for analyzing received signals, such as an electronic card or electronic processor, or a device such as a computer.

[0072] The non-cereal fruit sorting apparatus according to the present invention includes in particular a separation system for said fruits which intervenes according to feedback signals, which may take the form of a means of supplying compressed air or an articulated arm allowing the expulsion of said undesirable fruits.

[0073] Thus, the sorting device according to the present invention implements the non-destructive method of determining the maturity and quality of a non-cereal fruit according to the present invention.

[0074] Furthermore, the sorting apparatus according to the present invention is adapted to the size of the fruits being sorted. The size of these fruits can be, for example in the case of grapes, on the order of 10 to 20 mm in diameter, preferably 11 to 15 mm in diameter. • Analytical device, particularly portable:

[0075] In a particular embodiment, the object of the present invention can be applied to an analytical device, in particular a portable one, for analyzing plants in the field. For example, advantageously, the method for monitoring the state or development of a crop according to the present invention can be applied to a portable device according to FR2916850A1 or of an equivalent type.

[0076] Indeed, the device described in FR2916850A1, which, by its construction and method of use, is placed in contact with the plant, and in particular with the fruit, guarantees a constant excitation-target distance. The apparatus according to the present invention can therefore replicate the same characteristics as those described in FR2916850A1, adapted to expose the analyzed fruit(s) to the two wavelengths under the conditions described above.

[0077] In particular, such a (portable) device comprises a self-powered portable housing having a first face carrying user interface means and a second face called measurement face directed towards a measurement direction, this measurement face having on its periphery a surface carrying the excitation means and on its center a part extending in the measurement direction, containing at least a part of the detection electronics and carrying the detection means on its face on the side of the measurement direction.

[0078] In a particular embodiment, the (portable) device according to the present invention includes a positioning device in space, such as a GPS, which allows, for example, a precise link to be made between the measurements taken and a crop or part of a crop.

[0079] Furthermore, in the context of the present invention, the radiation systems of a (portable) device according to the present invention comprise one or more light-emitting diode (LED) sources. The advantage of LEDs is their reduced energy consumption.

[0080] In a particular embodiment, such a device includes at least one excitation emitter fixed on a heating element whose shape determines the position and orientation of said emitter or detector.

[0081] In particular, the handheld device includes management means arranged to provide at least one fluorescence measurement, for example, radiation at 650nm - referred to as "excitation", and at least one detection excitation (for example, at 750nm, used in reflection) determined to be combined and correlated with each other, as described herein.

[0082] Preferably, the management (and processing) resources are arranged to: control the pulsed radiation system(s), detect the fluorescence peaks generated by these pulses using detectors, which can be amplified to achieve the rejection of ambient light by a feedback loop, and process the detected fluorescence signal and provide the fluorescence measurement.

[0083] Preferably, the management (and processing) resources are arranged to: control the pulse radiation system(s), detect the reflection peaks generated by these pulses using detectors, which can be amplified to achieve the rejection of ambient light by a feedback loop, and process the detected reflection signal.

[0084] Preferably, the management and processing resources are arranged to: control the radiation system(s) according to a frequency including modulation, and process the fluorescence detection signal, for example by phase demodulation in order to obtain rejection of ambient light, and provide the fluorescence measurement.

[0085] Preferably, the management and processing resources are arranged to: control the radiation system(s) according to a frequency including modulation, process the reflection detection signal, for example by phase demodulation in order to obtain the rejection of ambient light, and provide the reflection measurement.

[0086] Thus, the device (in particular portable) according to the present invention, allows the implementation of a method configured for monitoring the evolution of a crop over time, characterized in that it allows the comparison of a plurality of measurements taken over time for a specific compound in the same plant or on the same plot.

[0087] Preferably, the method is implemented to assess the ripeness, quality, or composition of fruits or vegetables by measuring the content of at least one maturity marker, such as an anthocyanin.

[0088] Preferably, the method is implemented so that at least one measurement of a plurality of plants is carried out on the fly, as the measuring device (in particular portable) is moved according to the present invention. FIGURES

[0089] The following are examples of embodiments of the present invention, by way of non-limiting illustration, with reference to the accompanying figures in which: [ Fig.1 ] is a schematic representation of the sorting process according to the present invention applied to a non-cereal fruit sorting line; [ Fig. 2 ] represents a phase diagram for grape exposure.

[0090] With reference to the figure 1 , grapes 1 are put into scrolling (scrolling represented by arrow A) by means of a conveyor 2, the scrolling speed of the conveyor can eject the berries at its end along a known trajectory, or the grape 1 can be analyzed on the conveyor 2 (not shown here).

[0091] A first electromagnetic wave 3 is focused on the berries as they pass in front of the camera 4. The first electromagnetic wave is produced by a first emitter 5. A second electromagnetic wave 6 is focused on the berries as they pass in front of the camera 4. The second electromagnetic wave is produced by a second emitter 7. The return signals 8 are thus captured by the camera 4. Thus, to implement the method of measuring emitted fluorescence (750 nm) and reflection (750 nm), the emitters 5, 7 are turned on sequentially so that the camera 4 successively measures the fluorescence due to the excitation of each light source and the reflection signal.

[0092] Thus, camera 4 can in this particular case be represented as a monochrome infrared camera equipped with a 750 nm bandpass filter.

[0093] The lens of camera 4 can be equipped with a bandpass filter to specialize it on the wavelength of the fluorescence being sought.

[0094] The camera 4's line of sight can be perpendicular to the plane of the conveyor 2 which moves the grape 1 (represented by arrow A).

[0095] The camera sensor 4 can be oriented lengthwise so as to be transverse to the conveyor's travel axis 2.

[0096] The camera lens 4 can allow the camera 4's sharpness to be adjusted at the focal length.

[0097] One of the emitters 5,7 can be low-pass filtered to cut off fluorescence-specific electromagnetic waves and thus be totally dedicated to providing a reflection return signal 8.

[0098] The emitters 5,7 can be sized to illuminate all the bays 1 moving across the width of the conveyor 2. The camera 4 can be fitted with a lens so as to receive return signals 8 across the entire width of the conveyor 2.

[0099] The focal length and exposure intensity are sized to position the 5.7 emitters at the correct distance from the grape.

[0100] The 5.7 emitters can include LEDs to provide the desired electromagnetic waves. These LEDs can be arranged to achieve exposure in a continuous line on one side of the exposed fruit.

[0101] The exposure(s) can be adjusted in orientation to converge at a single point on the known trajectory of the non-cereal fruit.

[0102] Transmitters 5,7 and camera 4 can be sufficiently protected from radiation from the outside environment so as not to affect the measurements.

[0103] The emitters 5.7 can be configured to illuminate at least the width of conveyor 2 which carries the fruit to be sorted.

[0104] Transmitters 5 and 7 can be above conveyor 2 or shortly after the exit of conveyor 2.

[0105] Transmitters 5 and 7 can be offset so as not to be in the field of view of camera 4.

[0106] An ejection means 9, such as a nozzle, allows for the removal of unwanted fruit or objects. Indeed, an optional advantage of the present invention is the ability to identify unwanted objects, if desired, by comparing the image obtained after the fruit is exposed to the camera 4 with the expected general outline of the fruit to be sorted. Other sorting systems, based for example on color or shape, can also be added to increase sorting accuracy.

[0107] Here, it is an air jet 10 which allows the ejection of the unwanted fruit 11.

[0108] There are as many ejection means as necessary for sorting to be operational across the width of the conveyor.

[0109] Thus, the 12 sorted fruits that we wish to keep are isolated and recovered.

[0110] A management system is implemented to manage the lighting sequencing and fluorescence measurement synchronization. Thus, in figure 2 is represented a phase diagram particularly well suited to the sorting chain according to the figure 1 in which:

[0111] P1 represents excitation period No. 1. P1 can be used to define the resolution of the system; P2 represents an excitation period No. 2; T1 represents an excitation duration No. 1; T2 represents an excitation duration No. 2, the most sensitive (absorbed), can be adjusted to guarantee sufficient signal; T3 represents a time available for noise measurement; and D1 represents a phase shift to ensure synchronization of the reference and the measurement.

[0112] The excitation duration at the 625 nm wavelength is between 100 and 300 µS, preferably 240 µS. The total cycle time is between 300 and 600 µS, preferably 500 µS.

[0113] The excitation sequencing must allow for measurements at multiple locations on the same fruit being sorted (here, grapes) as it moves along at a known speed. The number of measurements taken on an average-sized piece of fruit defines the system's resolution. This resolution is determined by the required level of accuracy and the desired fruit throughput.

[0114] The duration of exposure in wavelengths and their periods can be managed by an automated system adapted for industrial vision.

[0115] The phase shift between several exposure wavelengths can also be managed by the controller.

[0116] The result of the processing is rendered in a simple form interpretable by the operator so as to allow him to adjust the selection to be made.

[0117] A juxtaposition of images from emitted fluorescence and images of reflection makes it possible to identify and quantify berries with low chlorophyll levels.

[0118] The device may provide for the analysis of images for elements with low chlorophyll levels in order to discriminate berries from other elements of the harvest.

[0119] It is advantageous, then, for the management system to control the sorting equipment in such a way as to carry out the selection set by the operator. This setting can be modified at any time during operation.

[0120] The operator may determine a selection threshold beforehand through appropriate analysis, either by first preparing a sample analyzed by other means, or by assessing the quantitative distribution between several batches in progress and by adjusting the selection threshold. EXAMPLES

[0121] As an illustration, absorbance measurements according to the present invention were carried out by spectrophotometry at 520 (i.e. a wavelength consistent with anthocyanin content) on grapes.

[0122] Thus, differences between riper and less ripe batches could be measured using the method according to the present invention: Table 1: Results obtained in 2020 Grape variety Percentage difference between the ripest and least ripe fruits Merlot +25,5% Cabernet 1 +13,6% Cabernet 2 +12,6%

[0123] Note: The 2020 process differed from the 2021 process in that the fruit classification was done manually based on the results provided by the process. Each fruit was manually passed through the system one by one, where it was excited at the appropriate wavelength. The order of passage was maintained for each fruit, and the process was manually applied to each fruit while viewing the recorded fluorescence images. The fluorescence resulting from each excitation was recorded separately. Table 2: Results obtained in 2021 Grape variety Percentage difference between the ripest and least ripe fruits Merlot 1 +48% Merlot 2 +37% Cabernet 1 +58%

[0124] The 2021 process showed a better overall level of maturity on the tested grapes compared to those of 2020.

[0125] The 2021 measuring bench was fully automated up to the physical classification of the fruits according to the process.

[0126] Analyses of anthocyanin content in finished wine from grapes classified as such during the 2021 trials show the same trends

Claims

1. Non-destructive method for determining the maturity and quality of a non-cereal fruit comprising at least two steps of exposing said fruit to at least two different electromagnetic emissions and of measuring their feedback signals, in which: - one of the steps of exposing said fruit comprises delivery of at least one wavelength capable of provoking a chlorophyll-fluorescence signal and measurement, where appropriate, of said chlorophyll-fluorescence signal; and - one of the steps of exposing said fruit to at least one electromagnetic emission comprises delivery of at least one selected wavelength capable of generating a reflection signal due to reflection from said fruit, and measurement, where appropriate, of said reflection signal, characterized in that said at least one wavelength capable of provoking fluorescence of chlorophyll is between 590 nm and 660 nm, and in that the measurements of the fluorescence signal of the chlorophyll and the reflection signal are taken by the same sensor.

2. Method according to Claim 1, characterized in that the sensor is a camera that is optionally provided with a band-pass filter, for example a band-pass filter calibrated at 750 nm.

3. Method according to either of Claims 1 and 2, characterized in that, on the one hand, the delivery of at least one wavelength capable of provoking a chlorophyll-fluorescence signal, and on the other hand, the delivery of at least one selected wavelength capable of generating a reflection signal due to reflection from said fruit, are sequenced and / or synchronized, with optionally at least one pause to allow measurement of background noise.

4. Method according to any of Claims 1 to 3, characterized in that said at least one wavelength selected to generate a reflection from said fruit is selected with a wavelength significantly longer than the wavelength of excitation of the chlorophyll.

5. Method according to Claim 3, characterized in that the wavelength significantly longer than the wavelength of excitation of the chlorophyll is between 700 nm and 900 nm, and preferably 750 nm plus or minus 10 nm, and in that said at least one wavelength capable of provoking fluorescence of chlorophyll is 625 nm plus or minus 10 nm.

6. Method according to any of Claims 1 to 5, characterized in that the non-cereal fruit contains at least one pigment selected from lycopene, beta-carotene, anthocyanin and lutein, preferably anthocyanin and / or lycopene.

7. Method according to any of Claims 1 to 6, characterized in that the non-cereal fruit is selected from a tomato, grape, strawberry, raspberry, redcurrant, cherry, blackberry, lingonberry, blackcurrant, aronia berry, cranberry, elderberry and blueberry, but is preferably a grape.

8. Apparatus for sorting non-cereal fruits comprising: - a device for supplying non-cereal fruits, - optionally at least one means for discarding objects that are not fruits; - at least one emitting system configured to generate at least one electromagnetic emission; - at least one detection region configured to measure and optionally analyse the feedback signals generated by said at least one emitting system; - a system for separating said fruits, which is activatable depending on the feedback signals, characterized in that: - said at least one emitting system is configured to expose said fruit to at least two different electromagnetic emissions; - one of said at least two electromagnetic emissions comprises at least one wavelength capable of provoking fluorescence of chlorophyll; - one of said at least two electromagnetic emissions comprises at least one wavelength selected so as to generate a reflection from said fruit, such as 750 nm; - said at least one emitting system and the detection region are arranged in such a way: - that at least two feedback signals generated by said at least one emitting system are picked up in the same detection region, or - as to be on the same side of the fruits characterized in that said at least one wavelength capable of provoking fluorescence of chlorophyll is between 590 nm and 660 nm, and in that the measurements of the fluorescence signal of the chlorophyll and the reflection signal are taken by the same sensor.

9. Method for sorting non-cereal fruits comprising the following steps: - bringing each non-cereal fruit individually into at least one exposure region, - exposing said fruit in said at least one exposure region to at least two different electromagnetic emissions, - receiving and analysing the feedback signals of the exposing step, - separating said fruits into different classes depending on their individual signals, in which: - one of the at least two different electromagnetic emissions comprises at least one wavelength capable of provoking fluorescence of chlorophyll by way of feedback signal; - one of the at least two different electromagnetic emissions comprises at least one wavelength selected so as to generate a reflection from said fruit by way of feedback signal; and - the values that define the classes are preset or selected depending on a distribution of the chlorophyll-fluorescence and reflection signals emitted by a sample of non-cereal fruits possessing known properties characterized in that said at least one wavelength capable of provoking fluorescence of chlorophyll is between 590 nm and 660 nm, and in that the measurements of the fluorescence signal of the chlorophyll and the reflection signal are taken by the same sensor.