System for assisting in the artificial insemination of animals, and method for using such a system

Abstract

The invention relates in particular to a system for assisting in the artificial insemination of an animal, comprising a vaginal penetration device, provided with an image sensor, and a data processing unit, the system being configured for: - obtaining (500) at least one image of a view close to the cervix of the animal; - determining (515-1, 515-2, 515-n) a plurality of features of elements represented in said at least one image; - estimating (520) a fertility index of the animal on the basis of said plurality of features; and - transmitting (525) the estimated fertility index to a user of the vaginal penetration device.

Description

[0001] System for assisting in the artificial insemination of animals and method of using such a system

[0002] TECHNICAL FIELD OF THE INVENTION

[0003]

[0001] The present invention relates to a system for assisting the artificial insemination of animals, making it possible to improve the efficiency of depositing semen in the female genital tract, particularly in terms of time and costs.

[0004]

[0002] The invention also relates to a method of using such a system and / or a method of assisting in the artificial insemination of animals.

[0005] STATE OF THE ART

[0006]

[0003] In order to carry out artificial insemination of animals such as cattle, pigs, equines, sheep or goats, a known device, presented in particular in document FR-A 1 -2690072, comprises an elongated gun having a front end allowing the operator to penetrate the vagina to reach the cervix and deposit in this uterus, as close as possible to the fallopian tubes, the liquid semen contained in a straw or in a capsule.

[0007]

[0004] However, for certain animals, in particular for cattle such as heifers which have never calved, passage through the vagina is difficult to achieve because it has numerous tight folds which are difficult to pass through and the entrance to the cervix is ​​difficult to locate, as well as to pass through, because it is very narrow.

[0008]

[0005] There is then a greater chance of making a seed deposit without result, which causes wasted time and wastes seeds presenting a certain cost.

[0009]

[0006] To overcome these difficulties, there are devices for vaginal penetration provided with video viewing means such as an endoscopy-type tube and having means for transmitting images to a remote screen, as described in document WO 2016 / 066962 and in document WO 2019 / 122685.

[0010]

[0007] Such devices make it possible in particular to facilitate the insertion of the end of a guide tube, called a speculum tube, between folds or narrow areas, in order to position in particular the tip of the gun at the precise location ensuring the best results of the operation.

[0008] Although such devices are satisfactory, there is a constant need to improve them to improve the efficiency of a semen deposit, particularly in terms of time and costs.

[0011] STATEMENT OF THE INVENTION

[0012]

[0009] The present invention aims to remedy all or part of the drawbacks of the state of the art cited above.

[0010] To this end, the invention aims, according to one aspect, at a system for assisting in the artificial insemination of an animal comprising a vaginal penetration device provided with an image sensor, and a data processing unit, the system being configured to:

[0013]

[0011] obtaining at least one image of a close view of the cervix of the animal;

[0014]

[0012] determining a plurality of characteristics of elements represented in said at least one image;

[0015]

[0013] estimating a fertility index of the animal based on said plurality of characteristics; and

[0016]

[0014] transmitting to a user of the vaginal penetration device the estimated fertility index.

[0017]

[0015] The system according to the invention thus makes it possible to help a user decide whether a semen deposit can be carried out with a chance of success, improving the artificial insemination process, particularly in terms of time and costs.

[0018]

[0016] The plurality of features may include at least one color feature and one geometry feature of an element contained in the at least one image.

[0017] The plurality of features may include an indication of an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the animal's cervix, a vaginal discharge, a presence of at least one cyst, and / or a presence of pus.

[0019]

[0018] The system may be further configured to deposit animal semen.

[0020]

[0019] The plurality of characteristics may be estimated for a plurality of images comprising the at least one image, the fertility index of the animal being estimated based on the plurality of characteristics estimated on the plurality of images.

[0021]

[0020] The system may be further configured to identify a cervix of the animal and determine the plurality of characteristics and estimate the fertility index after identifying the cervix.

[0022]

[0021] The system may be further configured to determine the plurality of features based on a convolutional neural network.

[0023]

[0022] The system may comprise a first device configured to obtain the at least one image and transmit to a user the estimated fertility index, and a second device distinct from the first device configured to determine the plurality of characteristics and estimate the fertility index.

[0024]

[0023] According to another aspect, the invention relates to a method for assisting in the artificial insemination of an animal using a system for assisting in the artificial insemination of an animal comprising a vaginal penetration device provided with an image sensor, and a data processing unit, the method comprising: obtaining at least one image of a close view of the cervix of the animal; determining a plurality of characteristics of elements represented in said at least one image; estimating a fertility index of the animal on the basis of said plurality of characteristics; and transmitting the estimated fertility index to a user of the vaginal penetration device.

[0025]

[0024] The method according to the invention thus makes it possible to help a user decide whether a semen deposit can be carried out with a chance of success, improving the artificial insemination process, particularly in terms of time and costs.

[0026]

[0025] The plurality of characteristics comprises for example at least one color characteristic and a geometry characteristic of an element contained in the at least one image. It may comprise an indication relating to an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the cervix of the uterus of the animal, a vaginal discharge, a presence of at least one cyst and / or a presence of pus.

[0027]

[0026] According to particular embodiments, the method further comprises a seed deposit.

[0028]

[0027] Still according to particular embodiments, the plurality of characteristics is estimated for a plurality of images comprising said at least one image, the fertility index of the animal being estimated on the basis of the plurality of characteristics estimated on the plurality of images, thus making it possible to improve the reliability of the results provided.

[0029]

[0028] Still according to particular embodiments, the method further comprises an identification of the cervix of the animal, the determination of a plurality of characteristics and the estimation of a fertility index being carried out only after identification of the cervix. Such embodiments make it possible to improve the comfort of the user by directly providing him with usable results.

[0030]

[0029] The determination of a plurality of characteristics is for example based on a convolutional neural network.

[0031]

[0030] According to particular embodiments, the steps of obtaining at least one image and transmitting to a user the estimated fertility index and the steps of determining a plurality of characteristics and estimating a fertility index are implemented in separate devices.

[0032]

[0031] The invention also relates to a device configured to implement the method described above or a part thereof. The advantages provided by this device are similar to those described previously with regard to the method.

[0032] A computer program, implementing all or part of the method described above, installed on pre-existing equipment, is in itself advantageous, since it helps a user of a vaginal penetration device used for artificial insemination to decide whether or not to deposit the semen.

[0033]

[0033] Thus, the present invention also relates to a computer program comprising instructions for implementing the method described above, in whole or in part, when this program is executed by a processor.

[0034]

[0034] This program may use any programming language (for example, an object language or other) and be in the form of interpretable source code, partially compiled code or fully compiled code.

[0035]

[0035] Another aspect relates to a non-transitory storage medium for a computer-executable program, comprising a data set representing one or more programs, said one or more programs comprising instructions for, upon execution of said one or more programs by a computer comprising a processing unit operatively coupled to memory means and to an input / output interface module, to execute all or part of the method described above.

[0036] BRIEF DESCRIPTION OF THE FIGURES

[0037]

[0036] Other advantages, aims and particular characteristics of the present invention will emerge from the following non-limiting description of at least one particular embodiment of the devices and methods which are the subject of the present invention, with reference to the appended drawings.

[0038]

[0037] Figures 1 a and 1 b are partial views of a system for assisting in the artificial insemination of animals, formed here by a device for vaginal penetration, comprising an insemination gun and an image sensor.

[0039]

[0038] Figure 2 is a detail view of the vision tube.

[0040]

[0039] Figure 3 illustrates an example of such a system for assisting in the artificial insemination of animals, for estimating a fertility index from images of the cervix.

[0041]

[0040] Figure 4 illustrates an example of steps for building a learning database and generating one or more numerical models for estimating a fertility index.

[0042]

[0041] Figure 5 illustrates an example of steps for estimating a fertility index based on one or more images of an animal's cervix and one or more digital models.

[0043]

[0042] Figures 6, 7 and 8 illustrate, schematically, examples of images of cervixes and indicate a fertility index estimated from these images.

[0043] Figure 9 illustrates an example of a device that can be used to implement, at least partially, embodiments of the invention, in particular steps described with reference to Figures 4 and 5.

[0044] DETAILED DESCRIPTION OF THE INVENTION

[0045]

[0044] The present description is given as a non-limiting example of embodiment.

[0046]

[0045] The inventors have observed that images, in particular images of cervixes, acquired using a device for vaginal penetration and provided with a video sensor or an image sensor, can be analyzed to estimate a fertility index that can be used by a user of the device to determine whether the conditions are met with a sufficient level of confidence to deposit a semen. Thus, if the conditions are not met, the user can postpone or abandon the deposit, saving straws and time.

[0047]

[0046] Figures 1 a and 1 b partially show a system for assisting in the artificial insemination of animals, formed here of a device 1 for vaginal penetration forming an assembly comprising an elongated shape which is substantially cylindrical, having towards the rear indicated by the arrow marked AR, a handle 2 receiving along its axis and sliding relative to the handle, a deposit gun 4 actuated by a push button 6.

[0048]

[0047] The handle 2 is extended forward by a guide tube 8, called a speculum tube, for example made of transparent plastic, which is connected to the handle 2 by means of an assembly ring 10. The guide tube 8 comprises inside an axial tube 12 surrounding an extension of the gun which is configured to cooperate with the insemination straw, and next to it a video vision tip 14 of an image acquisition system. The diameter of this guide tube 8 is in dimensions which respect the well-being of the animal and facilitate penetration.

[0049]

[0048] Figure 2 shows the video viewing tip 14 comprising a rear portion fitting into a front portion of the handle 2.

[0050]

[0049] The video viewing endpiece 14 may comprise, at the front, a light source, for example one or more cells of the LED type (acronym light-emitting diode in English terminology).

[0051]

[0050] The video vision tip 14 can be bent at a front end, for example over a length L and at an angle a, so as to protect it and in particular to obtain focusing of the image on the axis of the device at a desired working distance.

[0052]

[0051] The handle 2 may include a button for turning on the video system, which is not shown in the figures. The rear cell of the video system may communicate with a remote local device provided with display means such as a screen, which is separate from the device for penetration, by a wired connection, for example using a connection of the USB type (acronym for universal serial bus in English terminology) located on the handle, or by a wireless connection, for example using radio communication of the Bluetooth or WiFi type (Bluetooth and WiFi are trademarks). The rear cell may also carry out specific pre-processing, for example data normalization and / or conversion of the data into a particular format, for example a standard format.

[0053]

[0052] In the case of a wireless connection, the handle 2 contains the electronics for controlling the lighting and vision means, as well as an autonomous electrical power source.

[0054]

[0053] The remote local device is for example of the smartphone or tablet type. It can also be a terminal or a computer. It is provided with a communication interface compatible with that of the penetration device for receiving data, in particular raw data or images. These received data or images can be processed in an application of the remote local device, displayed, and / or transmitted to another device, for example a remote device for storage or processing purposes, using the communication interface compatible with that of the penetration device or using another communication interface, wired or wireless.

[0055]

[0054] The display system may be arranged for example in a waterproof cover comprising fixing means, in particular a strip comprising a quick-release fastener of the Velcro type (Velcro is a trademark) to fix it to the user's arm. It may also use any other vision system, in particular a system integrated into glasses with radio communication of the Bluetooth type.

[0056]

[0055] Such a device thus offers the advantage of monitoring the progress of the guide tube for more precise insemination.

[0057]

[0056] In fact, the operator can follow the progress of the guide tube on a stable and clean screen.

[0058]

[0057] In the example described, additional information may also be displayed, for example a fertility index estimated by analysis of the acquired images.

[0059]

[0058] Figure 3 illustrates an example of the animal artificial insemination assistance system configured to estimate a fertility index from images of cervixes.

[0060]

[0059] As illustrated, the system here comprises an apparatus for vaginal penetration provided with a vision (or image acquisition) system, for example the apparatus 1 illustrated in FIG. 1, a remote local device 300 for processing and displaying data and a remote storage and calculation device 305.

[0061]

[0060] The remote local device 300 for processing and displaying data is for example a smartphone provided with a WiFi or Bluetooth type communication interface for exchanging data with the device 1 and a cellular type communication interface (eg, 3G, 4G, 5G or similar) for exchanging data with the remote storage and calculation device 305 through a communication network 310, for example an Internet type communication network.

[0062]

[0061] According to particular embodiments, the remote local data processing and display device 300 receives raw data, for example raw images, from the apparatus 1. These raw data are analyzed by the remote local data processing and display device 300 to determine, preferably in real time, a fertility index using an algorithm and / or one or more digital models comprising parameters previously calculated in the remote storage and calculation device 305 and received from this device. The analysis can be carried out image by image or by sets of images.

[0063]

[0062] According to other embodiments, the raw data received by the remote local data processing and display device 300 are transmitted to the remote storage and calculation device 305 where they are analyzed to determine a fertility index. The latter is then transmitted to the remote local data processing and display device 300, with, optionally, other data, to be displayed. It is observed here that the raw data received by the remote local data processing and display device 300 can be processed by the latter before being transmitted to the remote storage and calculation device 305.

[0064]

[0063] The remote storage and calculation device 305 is for example a server provided with a network type communication interface for exchanging data with other devices through a communication network 310, for example an Internet type communication network.

[0065]

[0064] According to particular embodiments, the remote storage and calculation device 305 allows the processing of raw and / or pre-processed data and the learning of a digital model from previously stored data, to estimate a fertility index, as described with reference to figures 4 and 5.

[0066]

[0065] According to some embodiments, the remote storage and computing device 305 comprises one or more computing servers and one or more storage servers.

[0067]

[0066] Figure 4 illustrates an example of steps for constituting a learning database and generating one or more digital models for estimating a fertility index. These steps are, for example, implemented in the remote storage and calculation device 305 of Figure 3.

[0068]

[0067] As illustrated, a first step here has the object of obtaining one or more images (step 400). These images are for example obtained from the vaginal penetration device 1, using the remote local data processing and display device 300 which, on command from a user of the device, selects the images to be used and transmits them to the remote storage and calculation device 305 where they are advantageously stored in a database 405 to be processed.

[0069]

[0068] These images (or some of them) can then be accessed by a user who can determine characteristics of the elements represented on these images (steps 410-1 to 410-n), for example the following characteristics: the absence of blood, for example using the value 1 in the absence of blood and the value 0 in the presence of blood; the presence of vaginal folds on the cervix or cervical folds (forming a star-shaped geometry of the cervix), for example using the value 1 in the case of well-defined, non-edematous and homogeneous folds, and the value 0 in the case of less pronounced, less defined and edematous folds; the absence of prolapse, for example using the value 1 in the absence of prolapse and the value 0 in the presence of prolapse; a level of opening of the animal's cervix, for example on a scale from 0 to 1 (eg, 0 for a closed cervix and 1 for an open cervix); a level of vaginal discharge, for example, the value 0 in the absence of discharge (without color), the value 1 in the case of heat discharge (light pink color, translucent) and the value 2 in the case of pus discharge (red color, opaque); the presence of cyst in the vagina or on the cervix, for example, using the value 0 in the presence of a cyst and the value 1 in the absence of a cyst; the presence of abnormalities such as double cervix, vaginal band, hematomas, petechiae, adhesions, urovagina, scar and / or placental remains, for example, using the value 0 in the absence of these abnormalities and the value 1 in the presence of one or more of these abnormalities; and / or the absence of lochia, for example using the value 1 if lochia is absent and the value 0 if lochia is present.

[0070]

[0069] These characteristics are stored here in connection with the corresponding images, for example in the database 405.

[0071]

[0070] Other characteristics can of course be determined and stored, such as the presence of more pus (often characterized by a dark color).

[0072]

[0071] General characteristics, specific to the animal considered (at the time of taking images and / or their analysis), can also be identified and stored, for example the presence of pathology.

[0073]

[0072] In a following step, if a semen deposit was carried out at the time of taking the images considered, physiological measurements and / or observations are carried out and stored to characterize a state of fertility, on the day of taking the images, relating to the determined characteristics (step 415). This may be, for example, a fertility indicator on the day of taking the image and the semen deposit, the value of this fertility indicator being determined a posteriori, for example 30 days after taking the images and the semen deposit (for example the value 1 in the case of fertilization and the value 0 otherwise). It may also be the presence or absence of a fetus 60 days after taking the images and the semen deposit (for example the value 1 if a fetus is present and the value 0 otherwise).

[0074]

[0073] Then, if the quantity of data is sufficient (step 420), for example if the number of images stored and associated with characteristics and / or fertility observations is greater than a determined threshold, it may be decided to determine or update one or more digital models (step 430), for example one or more digital models for determining characteristics from images and / or a digital model for determining a fertility index from determined characteristics.

[0075]

[0074] As illustrated, this step of determining or updating one or more digital models may be preceded by preprocessing of the images (step 425). Such preprocessing consists, for example, of reducing the size of the images according to a single dimension, for example a dimension of 200 x 200 points (or pixels). It may also comprise normalizing the value of the points, for example to express the value of each point of each image, for each of the components (for example the components known as RGB), in a predetermined range, for example between the values ​​0 and 1. It is observed that the preprocessing applied here is preferably the same as that applied to the images from which a fertility index is estimated (as described with reference to FIG. 5.

[0076]

[0075] By way of illustration and depending on the type of digital models used, the number of images used to determine these digital models may be between several hundred images and several tens of thousands of images, for example ten thousand images.

[0076] The digital model(s) used to determine, from images of cervixes, characteristics such as an indication relating to an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the animal's cervix, a vaginal discharge, etc., may be artificial intelligence models, for example deep learning neural networks such as convolutional neural networks, also known as CNN (acronym for convolutional neural network in English terminology), for example convolutional neural networks known as MobileNet.

[0077]

[0077] The fertility index can be estimated in several ways. In particular, it can correspond to a linear combination of the values ​​of each characteristic with predetermined weights. It can also be estimated using a neural network, for example of the "Multilayer Perceptron" type, which will have been trained on learning data.

[0078]

[0078] The MobileNet network is a CNN adapted to image recognition, allowing the recognition and classification of objects, and which can be implemented in mobile devices for real-time applications. It is distinguished from a large part of CNNs by the convolution part. In a MobileNet network, the convolutions are split into two parts, with a deep convolution applied to each channel of the input image and a point convolution that combines the results of the deep convolutions. Artificial neural networks of the MobileNet type are described, for example, in the article entitled "MobilesNets: Efficient Convolutional Neural Networks for Mobile Vision Applications", A. Howard et al., April 17, 2017.Other types of neural networks that can be used to determine characteristics of features depicted in cervical images and / or estimate a fertility index, according to embodiments of the invention, are described in the document entitled “Review of deep learning: concepts, CNN architectures, challenges, application, future directions”, Laith Alzubaidi et al., Journal of Big Data, Springer Open, 2021.

[0079]

[0079] For resource reasons, the architecture of the digital model(s) used to estimate a fertility index may vary depending on whether it is implemented in a remote server-type device or in a remote local smartphone-type device. Furthermore, a single digital model may be used to determine several characteristics. Alternatively, a digital model may be associated with each characteristic to be determined.

[0080]

[0080] By way of illustration, the architecture of a convolutional neural network used, in a device with limited resources, to identify characteristics used to determine a fertility index, from images of the cervix, can be as follows: an input layer: 3 images (each image corresponding to one of the three basic colors of the same image) comprising for example 200 x 200 pixels;

[0081] - successive convolution and maxpooling layers (convolution followed by maxpooling, convolution and maxpooling layers), for example six blocks, to extract significant and discriminating features from the input images. The pooling layer provides a compact representation of the extracted features. The succession of convolution and pooling blocks allows for the identification of increasingly complex features; a dropout layer to regularize the learning of the neural network. This prevents overfitting to the training data and improves the generalization capacity of the neural network.This is a regularization layer used only during the training of the neural network (it is not used when the neural network is put into operation); a concatenation layer (or flatten layer) to transform the convolution results (also called maps) received in 2D (2 dimensions) into a vector (1 D, 1 dimension) of characteristics. This layer is used to bridge the gap between the convolution layers and the fully connected layers; and two interconnection layers (or dense layers) to o decode the input vector (1 D) corresponding to the characteristics extracted by the convolution blocks and grouping into an output vector (1 D) which corresponds to the final classification of the input image; and o provide the estimated value(s) (output layer).

[0082]

[0081] The training of such neural networks can be carried out using the data obtained in accordance with the steps described with reference to Figure 4, using a learning algorithm such as those described in the documents cited above. These digital models make it possible to associate images with characteristics such as the absence of blood, the presence of cervical folds, the absence of prolapse, an opening of the animal's cervix, vaginal discharge, the presence of cysts, the presence of pus, etc.

[0083]

[0082] A similar or different numerical model, for example a logistic regression model, can be used to associate characteristics such as an indication relating to an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the animal's cervix, a vaginal discharge, a presence of cysts, a presence of pus, etc., with a fertility index. The latter can represent a probability of fertilization in the event of semen deposition.

[0084]

[0083] Figure 5 illustrates an example of steps for estimating a fertility index on the basis of one or more images of an animal's cervix and one or more digital models. These steps are, for example, implemented in the remote local data processing and display device 300 or in the remote storage and calculation device 305 of Figure 3.

[0085]

[0084] As illustrated, a first step here has the object of obtaining one or more images (step 500). Like the images obtained in step 400 of FIG. 4, these images are for example obtained from the vaginal penetration device 1, using the remote local device 300 for processing and displaying data.

[0086]

[0085] Preprocessing is preferably applied to the obtained images (step 505). As described with reference to step 425 of FIG. 4, the preprocessing consists, for example, of reducing the size of the images according to a single dimension, for example a dimension of 200 x 200 points (or pixels) and / or of normalizing the value of the points, for example to express the value of each point of each image, for each of the components (for example the components known as RGB), in a predetermined range, for example between the values ​​0 and 1.

[0087]

[0086] A following, optional step is intended to determine whether the images obtained can be used to determine the desired characteristics (step 510). By way of illustration, this step may comprise a step of identifying a cervix. If a cervix is ​​not identified in the images, the latter are not analyzed, the algorithm then continues by obtaining other images (step 500). On the contrary, if the images can be used to determine the desired characteristics, they are preferably stored, for example in the database 405, so that they can be used later to improve the digital models, as described with reference to FIG. 4.

[0088]

[0087] In one or more subsequent steps (steps 515-1 to 515-n), the image or each of the images is analyzed to determine the characteristics sought, for example, an indication relating to an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the cervix of the animal, a vaginal discharge, a presence of cysts, a presence of pus, etc. For these purposes, the image or each of the images is used as input to the digital model(s) whose output(s) represent, for example, the value of each characteristic sought.

[0089]

[0088] According to particular embodiments, the obtained values ​​of an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the cervix of the animal, a vaginal discharge, a presence of cysts, a presence of pus, etc., represent a probability of absence of blood, a probability of presence of cervical folds, a probability of absence of prolapse, a probability of opening of the cervix, a probability of vaginal discharge, a probability of presence of cysts, a probability of presence of metritis, etc., respectively. Still according to particular embodiments, an obtained value of vaginal discharge represents a probability of absence of vaginal discharge, value 0, a probability that the vaginal discharge is a heat discharge (light pink color, translucent), value 1, and / or a probability that the vaginal discharge is a pus discharge (red color, opaque), value 2.

[0090]

[0089] When multiple images are analyzed, an average of the results obtained for each characteristic can be performed to determine an average level for each characteristic.

[0091]

[0090] The determined characteristics (or their average values) are then used to estimate a fertility index (step 520) using a digital model, for example a digital model determined as described with reference to Figure 4. This fertility index is then indicated to a user of the vaginal penetration device used for insemination in order to assist him in his decision whether or not to deposit the semen (step 525). This indication may in particular be visual, for example by a value and / or a color displayed on the remote local device, or audible. The characteristics determined and used to estimate the fertility index, or some of them, may also be indicated to the user.

[0092]

[0091] Thresholds can further be set to assist the user. Thus, for example, if the fertility index is higher than a first insemination threshold, for example set at 0.5, the fertility index can be displayed in green, if it is lower than this first insemination threshold, but higher than a second insemination threshold, for example set at 0.4, the fertility index can be displayed in orange and if it is lower than this second threshold, the fertility index can be displayed in red.

[0093]

[0092] As illustrated, the process can continue as long as images are obtained.

[0094]

[0093] Figures 6, 7 and 8 illustrate, schematically, examples of images of cervixes and indicate a fertility index estimated from these images. For each of the images the determined characteristics are indicated as well as the fertility index.

[0095]

[0094] For example, from the image shown in Figure 6, it is determined that the probability of no blood is 0.88 (the probability of blood being 0.12), the probability of cervical folds being 0.77 (the probability of no cervical folds being 0.23), the probability of no prolapse is 0.57 (the probability of prolapse being 0.43), the probability of the cervix being closed is 0.66 (the probability of it being open is 0.34), the probability of no vaginal discharge is 0.08, the probability of the vaginal discharge being a heat discharge is 0.86, and the probability of the vaginal discharge being a pus discharge is 0.06. From these characteristics, it is estimated that the fertility index is 0.58. This index may be displayed in green if it is considered particularly favorable.

[0096]

[0095] Still as an example, from the image shown in Figure 7, it is estimated that the fertility index is 0.46. This index can be displayed in orange if it is considered not to be particularly favorable, but nevertheless not especially unfavorable.

[0097]

[0096] Finally, from the image shown in Figure 8, it is estimated that the fertility index is 0.38. This index can be displayed in red, or even crossed out, if it is considered to be particularly unfavorable.

[0098]

[0097] Of course, there are many other ways of displaying this information or part of it.

[0098] Figure 9 illustrates an example of a device that can be used to implement, at least partially, embodiments of the invention, in particular steps described with reference to Figures 4 and 5.

[0099]

[0099] The device 900 is for example a server, a computer, a terminal or a personal device such as a smartphone or a tablet.

[0100]

[0100] The device 900 preferably comprises a communication bus 902 to which are connected: a central processing unit or microprocessor 904 (CPU, acronym for central processing unit in English terminology); a read-only memory 906 (ROM, acronym for read only memory in English terminology) which may comprise the operating system and programs such as "prog", "progl" and "prog2"; a random access memory or cache memory 908 (RAM, acronym for random access memory in English terminology) comprising registers adapted to record variables and parameters created and modified during the execution of the aforementioned programs; a communication interface 910 connected to a distributed communication network 912, for example a wireless communication network and / or a local communication network, the interface being capable of transmitting and receiving data, in particular to and from other devices;and a graphics card, a sound card, and / or an audio / video card 914 which can be connected to a screen and / or speakers 916.;

[0101]

[0101] Optionally, the device 900 may also have the following elements: a hard disk 918 which may contain the aforementioned programs "prog", "progl" and "prog2" and data processed or to be processed according to the invention; an input / output interface 920 to which a keyboard 922, a mouse 924 and / or any other pointing device such as an optical pen, a touch screen, a voice recognition device, a gesture recognition device, a camera, a microphone or a remote control may be connected allowing the user to interact with the programs according to the invention; and / or a reader 930 of removable storage media 932 such as a memory card.

[0102]

[0102] The communication bus allows communication and interoperability between the different elements included in the device 900 or connected to it. The representation of the bus is not limiting and, in particular, the central unit is capable of communicating instructions to any element of the device 900 directly or via another element of the device 900.

[0103] The executable code of each program enabling the programmable device to implement the processes according to the invention may be stored, for example, in the hard disk 918 or in read-only memory 906.

[0103]

[0104] According to a variant, the executable code of the programs may be received via the communication network 912, via the interface 910, to be stored in a manner identical to that described previously.

[0104]

[0105] More generally, the program(s) may be loaded into one of the storage means of the device 900 before being executed.

[0105]

[0106] The central unit 904 will control and direct the execution of the instructions or portions of software code of the program(s) according to the invention, instructions which are stored in the hard disk 918 or in the read-only memory 906 or in the other aforementioned storage elements. When the power is switched on, the program(s) which are stored in a non-volatile memory, for example the hard disk 918 or the read-only memory 906, are transferred into the random access memory 908 which then contains the executable code of the program(s) according to the invention, as well as registers for storing the variables and parameters necessary for implementing the invention.

[0106]

[0107] Depending on the embodiment selected, certain acts, actions, events, or functions of each of the methods described herein may be performed or occur in a different order than they were described, or may be added, merged, or may not be performed or occur, as the case may be. In addition, in some embodiments, certain acts, actions, or events are performed or occur concurrently and not successively.

[0107]

[0108] Although described through a number of detailed exemplary embodiments, the proposed method and the equipment for implementing the method include various variations, modifications and improvements which will be apparent to those skilled in the art, it being understood that these various variations, modifications and improvements are part of the scope of the invention, as defined by the claims which follow. In addition, different aspects and features described above may be implemented together, or separately, or substituted for each other, and all of the different combinations and sub-combinations of the aspects and features are part of the scope of the invention. Furthermore, some systems and equipment described above may not incorporate all of the modules and functions described for the preferred embodiments.

Claims

Claims 1. System for assisting in the artificial insemination of an animal comprising a vaginal penetration device provided with an image sensor, and a data processing unit, the system being configured to: obtain (500) at least one image of a close view of the cervix of the animal; determine 515-1, 515-2, 515-n) a plurality of characteristics of elements represented in said at least one image; estimate (520) a fertility index of the animal on the basis of said plurality of characteristics; and transmit (525) to a user of the vaginal penetration device the estimated fertility index.

2. The system of claim 1, wherein the plurality of characteristics comprises at least one color characteristic and one geometry characteristic of an element contained in the at least one image.

3. The system of claim 1, wherein the plurality of characteristics includes an indication of an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the animal's cervix, a vaginal discharge, a presence of at least one cyst and / or a presence of pus.

4. The system of any one of claims 1 to 3, further configured to deposit animal semen.

5. The system of any one of claims 1 to 4, wherein the plurality of characteristics is estimated for a plurality of images including said at least one image, the fertility index of the animal being estimated based on the plurality of characteristics estimated on the plurality of images.

6. The system of any one of claims 1 to 5, further configured to identify a cervix of the animal and determine the plurality of characteristics and estimate the fertility index after identifying the cervix.

7. The system of any one of claims 1 to 6, further configured to determine the plurality of features based on a convolutional neural network.

8. System according to any one of claims 1 to 7, comprising a first device configured to obtain the at least one image and transmit to a user the estimated fertility index, and a second device distinct from the first device configured to determine the plurality of characteristics and estimate the fertility index.

9. Method for assisting in the artificial insemination of an animal using a system for assisting in the artificial insemination of an animal comprising a vaginal penetration device provided with an image sensor, and a data processing unit, the method comprising: obtaining (500) at least one image of a close view of the cervix of the animal; determining (515-1, 515-2, 515-n) a plurality of characteristics of elements represented in said at least one image; estimating (520) a fertility index of the animal on the basis of said plurality of characteristics; and transmitting (525) to a user of the vaginal penetration device the estimated fertility index.

10. The method of claim 9, wherein the plurality of characteristics comprises at least one color characteristic and one geometry characteristic of an element contained in the at least one image.

11. The method of claim 9, wherein the plurality of characteristics comprises an indication of an absence of blood, a presence of cervical folds, an absence of prolapse, an opening of the cervix of the animal, a vaginal discharge, a presence of at least one cyst and / or a presence of pus.

12. A method according to any one of claims 9 to 11, further comprising a seed deposit.

13. A method according to any one of claims 9 to 12, wherein the plurality of characteristics is estimated for a plurality of images comprising said at least one image, the fertility index of the animal being estimated on the basis of the plurality of characteristics estimated on the plurality of images.

14. A method according to any one of claims 9 to 13, further comprising identifying the cervix of the animal, the determination of a plurality of characteristics and the estimation of a fertility index being carried out only after identification of the cervix.

15. A method according to any one of claims 9 to 14, wherein the determination of a plurality of characteristics is based on a convolutional neural network.

16. Method according to any one of claims 9 to 15, according to which the steps of obtaining at least one image and transmitting to a user the estimated fertility index and the steps of determining a plurality of characteristics and estimating a fertility index are implemented in separate devices.

17. Computer program comprising instructions for implementing each of the steps of the method according to any one of claims 9 to 16 when the computer program is implemented in a computer.

Images