System for identifying animals, computer-implemented method, computer program and non-volatile data carrier

Abstract

Individual animals in a herd are identified using a first camera (140) that registers 3D image data (Dimg) of a back portion of a preselected individual (130). A database (160) stores reference data records, which each is associated with the identity of a parti- cular individual. The 3D image data (Dimg) is processed to extract data (Dext) describing anatomy / topology of the back portion of the individual (130); the extracted data (Dext) are matched against the reference data records; and an identity is assigned to the prese- lected individual (130) in response to a match between the extrac- ted data (Dext) and one of the reference data records, where the assigned identity is the identity associated with the matching re- ference data record. The reference data record for an individual is updated repeatedly if the 3D image data (Dimg) is obtained toge- ther with an independent identification (ID) the individual (130). However, such updating occurs if and only if at least one updating criterion is fulfilled, which relates to a physiology parameter (P), a behavioral parameter (B), a reproductive-cycle parameter (R) and / or a lactation-cycle-related parameter (L) for the individual (130) in question.

Description

[0001] System for Identifying Animals, Computer-Implemented Method, Computer Program and Non-Volatile Data Carrier

[0002] TECHNICAL FIELD

[0003] The present invention relates generally to automatic identification of animals. Especially, the invention relates to a system according to the preamble of claim 1 and a corresponding computer-implemented method. The invention also relates to a computer program and a non-volatile data carrier storing such a computer program.

[0004] BACKGROUND

[0005] Inter alia for health and food quality reasons, today’s dairy industry requires that the extracted milk can be traced back to each animal from which the respective milk originates. This, in turn, demands an unfailing system for identifying the animals in connection with each milking. Further, in herd management, there are numerous of examples of situations when reliable animal identification is key, for example at feeding stations and when determining a body condition score (BCS) for an animal. Below follows prior-art examples of image-based systems that for example may be used to identify individual animals.

[0006] US 2010 / 0246970 discloses a device and method for providing information about animals walking through an animal passage. The information comprises at least the number of animals walking through the animal passage. A detection device is used, which has a sensor device connected to a processor for capturing animal data about animals walking through the animal passage. For the purpose of outputting counter impulses when animals are detected in said signals, an analysis device recognizes animals in the data / signals captured by the sensor device. The sensor device is designed for producing tree-dimensional (3D) images, and the analysis device is designed for detecting animals in the 3D data of the 3D images and for counting the animals using said detection.

[0007] WO 2017 / 030448 shows a method and apparatus for automatically evaluating an animal based on its physical appearance. The document also shows a method and apparatus for determining a body condition score (BCS) or lameness, particularly for a dairy cow. Further, methods for automatically evaluating a body condition score for cattle are described. The methods may involve receiving 3D shape information corresponding to a space occupied by the animal; creating a 3D point cloud from the 3D shape information; creating a 3D model based on the shape information; calculating one or more representative measurements from the 3D model; and evaluating the animal based on the measurements.

[0008] WO 2021 / 032890 reveals a rotary milking platform that comprises a plurality of stalls and a radio frequency identification (RFID) animal identifying system for identifying animals entering the stalls of the platform. A microprocessor reads signals from an image capturing device and computes a feature vector from the captured image of each animal. A plurality of reference feature vectors comprising respective matrices of metrics already derived from images of the respective animals captured by the image capturing device are stored and cross-referenced with the identity of the respective animals. The microprocessor compares computed feature vectors of each animal with the stored reference feature vectors until a best match has been determined with one of the reference feature vectors. The identity of the animal of that matching reference feature vector is then determined as the identity of the animal of that computed feature vector. The determined identity of the animal in the relevant stall is compared with the identity of the animal determined for that stall by the RFID system. On a favorable comparison the identity of the animal determined from the captured image of that animal is confirmed as the identity of the animal. In the event of a conflict between the two identities being determined, a conflict alert signal is produced.

[0009] US 1 1 ,080,522 describes a system and a method for identification of individual animals based on images, such as 3D images, of the animals, especially of cattle and cows. When animals live in areas or enclosures where they freely move around, it can be complicated to identify the individual animal. Inter alia, the disclosure relates to a method for determining the identity of an individual animal in a population of animals with known identity. The method comprises the steps of acquiring at least one image of the back of a preselected animal, extracting data from said at least one image relating to the anatomy of the back and / or topology of the back of the preselected animal, and comparing and / or matching said extracted data against reference data corresponding to the anatomy of the back and / or topology of the back of the animals with known identity, thereby identifying the preselected animal. The method and system can be used to monitor feed intake, such as feed intake for dairy cows as well as health status.

[0010] Thus, various image-based solutions are known for analyzing animal characteristics, for example for identification purposes. However, these solutions may be problematic due to the large amounts of data that must be processed. Handling such data quantities is either expensive, time consuming, or both.

[0011] SUMMARY

[0012] The object of the present invention is to offer a solution that mitigates the above problem, and thus allows efficient and reliable image-based identification of animals.

[0013] According to one aspect of the invention, the object is achieved by a system for identifying individuals in a herd of animals in which each individual has a known identity. The system includes a first camera, a database and a controller. The first camera is configured to obtain 3D image data representing a back portion of a preselected individual in the herd of animals. The database contains reference data records, wherein each reference data record is associated with the identity of a particular one of the individuals in the herd of animals. The controller is configured to process the 3D image data to extract data describing an anatomy and / or a topology of the back portion of the preselected individual. The controller is also configured to match the extracted data against the reference data records in a database, and assign an identity to the preselected individual in response to a match between the extracted data and one of the reference data records in the database. The assigned identity is here the identity that is associated with the matching reference data record in the database. Further, the controller is configured to update, repeatedly, each of the reference data records in the database. The updating involves obtaining 3D image data representing the back portion of one of the individuals in the herd of animals together with an identification of said one individual that is independent from the 3D image data. Specifically, the controller is configured to update the reference data record in the database if and only if at least one updating criterion is fulfilled for said one individual. The at least one updating criterion, in turn, relates to a physiology parameter, a behavioral parameter, a reproductive-cycle parameter and / or a lactation-cycle- related parameter for said one individual.

[0014] This system is advantageous because it ensures that the reference data records are only updated when needed due to an expected change in the respective individual’s visual appearance. In other words, unnecessary updatings are avoided and the overall data handling process becomes more efficient.

[0015] According to one embodiment of this aspect of the invention, the controller is configured to derive the lactation-cycle-related parameter based on at least one milk-related parameter for said one individual. Here, the at least one milk-related parameter reflects: an amount of milk produced by said one individual, a fat-compensated amount of milk produced for said one individual, a duration of a milking session of said one individual, a milk flow rate during a milking session for said one individual and / or a number of days in milk for said one individual. Thus, the reference data records may be updated based on the characteristics of various lactation- cycle-related parameters that typically are correlated with varia- tions in the anatomy and / or topology of the back portion of dairy animals.

[0016] According to another embodiment of this aspect of the invention, the at least one updating criterion is considered to be fulfilled if a slope of at least one first graph fulfills at least one steepness criterion, for instance relating to the slope exceeding or falling below a first or second steepness threshold respectively. The at least one first graph here describes the amount of milk produced by said one individual, the fat-compensated amount of milk produced for said one individual, the duration of a milking of said one individual and / or the milk flow rate during a milking session for said one individual respectively as a function of time in relation to a reference point-in-time for said one individual. The reference point-in- time may be related to a number of days-in-milk for the individual, i.e. a measure of how far into the lactation period the animal is. Consequently, highly relevant criteria may be selected for updating the reference data records.

[0017] According to yet another embodiment of this aspect of the invention, the behavioral parameter reflects an activity level of said one individual and at least one of the at least one updating criterion is considered to be fulfilled if an average value of the activity level of said one individual exceeds a threshold level during a predefined measurement period prior to a point in time when the 3D image data was registered. As a result, criteria for updating the reference data records may for example be based on an indicator that the animal is in heat and / or shows an unnatural behavior.

[0018] According to still another embodiment of this aspect of the invention, the physiology parameter reflects a BCS of said one individual and at least one of the at least one updating criterion is considered to be fulfilled if a slope of a second graph fulfills a steepness criterion. The second graph describes the BCS as a function of time in relation to a reference point-in-time for said one individual, where for example the reference point-in-time is defined with respect to the number of days in milk. As well known, the BCS is strongly correlated with the appearance of cattle. Therefore, this parameter is a useful basis for determining whether a reference data record should be updated.

[0019] According to another embodiment of this aspect of the invention, the extracted data includes a set of image features derived from the 3D image data, where the set of image features characterize the anatomy and / or the topology of the back portion of the preselected individual. Alternatively, or in addition thereto, the extracted data includes a compressed version of the 3D image data. In either case, the data may be stored on a compact format.

[0020] According to a further embodiment of this aspect of the invention, the system includes a radio transceiver system that is configured to read out a code from an RFID tag carried by said one individual. The code, in turn, forms a basis for the above-mentioned identification that is independent from the 3D image data. Alternatively, or in addition thereto, the system may include an imaging subsystem that is configured to register image data in the form of an optically readable code on a tag carried by said one individual. The optically readable code, which may be a barcode, a quickresponse (QR) code and / or conventional characters, here forms a basis for said identification being independent from the 3D image data. Hence, provided that the individual is identified via the RFID code or the optically readable code and the at least one updating criterion is fulfilled, its reference data record may be updated.

[0021] According to yet another embodiment of this aspect of the invention, the system includes at least one data interface, e.g. a wired interface or a radio interface, that is configured to obtain the physiology parameter, the behavioral parameter, the reproductive- cycle parameter and / or the lactation-cycle-related parameter for said one individual, for instance via one or more sensors in a milking system. As a result, it is straightforward for the controller to check whether the at least one updating criterion is fulfilled.

[0022] According to yet another aspect of the invention, the object is achieved by a computer-implemented method, which is performed in at least one processing unit in a controller, which controller, in turn, is included in the above-proposed system.

[0023] The method involves obtaining 3D image data representing a back portion of a preselected individual in the herd of animals; processing the 3D image data to extract data describing at least one of an anatomy and a topology of the back portion of the preselected individual; and matching the extracted data against reference data records in a database, wherein each reference data record is associated with the identity of a particular one of the individuals in the herd of animals. In response to a match between the extracted data and one of the reference data records in the database, the method further involves assigning an identity to the preselected individual, which assigned identity is the identity that is associated with the matching reference data record in the database. The method further involves updating, repeatedly, each of the reference data records in the database. The updating here comprises obtaining 3D image data that represent the back portion of one of the individuals in the herd of animals together with an identification of said one individual, which identification is independent from the 3D image data. Specifically, the reference data record in the database is updated if and only if at least one updating criterion is fulfilled for said one individual. The at least one updating criterion relates to a physiology parameter, a behavioral parameter, a reproductive-cycle parameter and / or a lactation-cycle-related parameter for said one individual. The advantages of this method, as well as the preferred embodiments thereof, are apparent from the discussion above with reference to the proposed system.

[0024] According to a further aspect of the invention, the object is achieved by a computer program loadable into a non-volatile data carrier communicatively connected to at least one processing unit. The computer program includes software for executing the above method when the program is run on the at least processing unit.

[0025] According to another aspect of the invention, the object is achie- ved by a non-volatile data carrier containing the above computer program.

[0026] Further advantages, beneficial features and applications of the present invention will be apparent from the following description and the dependent claims.

[0027] BRIEF DESCRIPTION OF THE DRAWINGS

[0028] The invention is now to be explained more closely by means of preferred embodiments, which are disclosed as examples, and with reference to the attached drawings.

[0029] Figure 1 schematically illustrates an animal identification system according to one embodiment of the invention;

[0030] Figure 2 shows an animal carrying various types of tags that may be used for identification according to embodiments of the invention;

[0031] Figures 3-6 show graphs that exemplify different kinds of parameters, which may be used according to embodiments of the invention for determining whether at least one updating criterion is fulfilled for updating a reference record for a particular individual; and

[0032] Figure 7 illustrates, by means of a flow diagram, the general method for identifying animals according to the invention.

[0033] DETAILED DESCRIPTION

[0034] Figure 1 shows system 100 according to one embodiment of the invention, which system 100 is arranged for identifying individuals in a herd of animals, where each individual has a known identity. This means that, for a particular individual in the herd of animals, the system 100 selects an identity to be assigned to that indivi- dual, which identity is selected from a predefined group of potential identities that represent all the individuals in the herd of animals.

[0035] Preferably, the identification is effected at a place where only one individual at the time may be located, for example in a physically delimited area at or in close proximity to a milking point. In the present disclosure, the animal that is located at such an identification place is defined as a “preselected individual,” which in Figure 1 is indicated by the reference numeral 130.

[0036] The system 100 contains a first camera 140, a database 160 and a controller 110. The first camera 140 is configured to obtain 3D image data Dimgrepresenting a back portion of the preselected individual 130 in the herd of animals.

[0037] The database 160 includes reference data records, wherein each reference data record is associated with the identity of a particular one of the individuals in the herd of animals. According to embodiments of the invention, each reference data record may contain a set of image features characterizing the anatomy and / or the topology of the back portion. The anatomy, may here be represented by various relative distances between different parts of the animal, such as its hips, tail head, tail head nadirs, foreribs, short rib starts, hook starts, hook anterior midpoints, hook posterior midpoints, hip joints / thurls, pins, tail head junctions, tail and / or posterior midpoints. The topology may here be represented by the specific curvatures and shapes of the surface of the animal’s back.

[0038] The controller 110 is configured to process the 3D image data Dimg to extract data Dext describing the anatomy and / or the topology of the back portion of the preselected individual 130. According to embodiments of the invention, the extracted data Dext includes a set of image features derived from the 3D image data Dimg, which set of image features characterize the anatomy and / or the topology of the back portion of the preselected individual 130, for example as exemplified above. Alternatively, or in addition thereto, the extracted data Dext may include a compressed version of the 3D image data Dimgitself.

[0039] The controller 110 is further configured to match the extracted data Dext against the reference data records in a database 160, and in response to a match between the extracted data D ext and one of the reference data records in the database 160, the controller 110 is configured to assign an identity to the preselected individual 130. The assigned identity is here the identity that is associated with the matching reference data record in the database 160.

[0040] Additionally, in order to adapt the above image-based identification to gradual changes in the respective appearances of the individuals in the herd of animals, the controller 110 is configured to update, repeatedly, each of the reference data records in the database 160. To make sure that any updated image data represents the correct individual, the updating is exclusively made together with an identification of the individual 130, which identification is independent from the 3D image data Dimg. According to embodiments of the invention, the system may include a radio transceiver system 150 and / or an imaging subsystem 170 to effect the independent identification of the individual 130 by reading out a code from an RFID tag 135 or an optically readable tag 245 respectively, which tag(s) 135 / 245 is / are carried by the individual 130.

[0041] The radio transceiver system 150 is configured to read out the code from the RFID tag 135 that is carried by the individual 130. The code read out from the RFID tag 135, in turn, forms a basis for the identification ID being independent from the 3D image data D img . In addition to an antenna 151 , the radio transceiver system 150 may contain a receiver and demodulation unit configured to convert radio signals received from the RFID tag 135 via a radio interface 155 into a first characterizing signal adapted to be received by the controller 110, which first characterizing signal reflects an identity ID of the individual 130. Analogously, the imaging subsystem 170 is configured to register image data representing an optically readable code on the tag 245, such as a registration number, which optically readable code, in turn, forms a basis for the identification ID being independent from the 3D image data Dimg. Of course, either of these processes may involve converting the read out code into another code before the individual 130 can be identified, for instance via a conversion table. The imaging subsystem 170 may contain a camera 171 and an imaging processing unit 173, which is configured to obtain image information from the camera 171 and extract a second characterizing signal from said image information, which second characterizing signal reflects the identity ID of the individual 130 and which second characterizing signal is adapted to be received by the controller 110.

[0042] Irrespective of how the identity ID is obtained, provided that the individual’s 130 identity ID can be confirmed independently from the 3D image data Dimg, the reference data record for the individual 130 may be updated in the database 160. This updating involves obtaining 3D image data Dimg representing the back portion of the individual 130. However, to ensure that the reference data record is not updated unnecessarily, i.e. when updating is not needed, for example because it cannot be expected that the individual’s 130 back portion appearance has changed in a distinguishable manner since a latest update, the controller 110 is configured to update the reference data record in the database 160 if and only if at least one updating criterion is fulfilled for the individual 130 in question. The at least one updating criterion relates to: a physiology parameter P, a behavioral parameter B, a reproductive-cycle parameter R and / or a lactation-cycle-related parameter L for the individual 130. The parameters P, B, R and L will be discussed below with reference to Figures 3 to 6.

[0043] Preferably, the system 100 has at least one data interface 180 that is configured to obtain at least one of the physiology parameter P, the behavioral parameter B, the reproductive-cycle parameter R and / or the lactation-cycle-related parameter L for each of the individuals 130 in the herd of animals. The at least one data interface 180 may be a wired interface and / or a radio interface, which is connected, directly or indirectly, to one or more sensor units and / or processing units that for example form part of a milking system. Thus, the controller 110 may conveniently check whether the at least one updating criterion is fulfilled.

[0044] Figure 2 shows an animal 130 carrying three different types of tags 135, 245 and 235 respectively, which tags may be used for identifying the individual 130 according to embodiments of the invention as described above.

[0045] The radio transceiver system 150 may include an RFID reader arrangement, e.g. or LF (low frequency) or UHF (ultrahigh frequency) type. LF RFID systems operate on frequencies from 30 kHz to 300 kHz, typically around 125 kHz, and normally have a range in the order of a few decimeters up to around one meter. UHF RFID systems operate on frequencies from 300 MHz to 3 GHz, typically between 860 MHz and 960 MHz, and normally have a range around 10 meters. Alternatively, the radio transceiver system 150 may include a UWB (ultrawide band) communication system, which, in principle, may have a line-of-sight range. In practice, however, the range seldom exceeds 100 meters. In the present invention, a substantially shorter range is desired, say in the order of a few decimeters, since preferably only one individual at the time should be able to be located in the area where identification is effected, i.e. to represent the preselected individual.

[0046] The RFID tag 135 is implemented in a technology suitable for the radio transceiver system 150. This means that if the radio transceiver system 150 contains an RFID system, the RFID tag 135 contains a standard RFID transceiver. Analogously, if the radio transceiver system 150 contains a UWB communication system, the RFID tag 135 contains a UWB transceiver.

[0047] Preferably, the imaging subsystem 170 contains a camera 171 that is arranged, i.e. has such location and field of view that the image data registered by the camera 171 covers a predefined portion of the individual 130, which predefined portion is expected to include the tag 245 with the optically readable code, which tag 245 is carried by the individual 130. For example, the predefined portion of the individual 130 may at least cover the individual’s 130 right ear.

[0048] As an alternative, or supplement to the above tags 135 and 245, the individual 130 may carry a motion sensor unit 235 that is configured to register movement patterns and / or behavioral patterns for the individual 130. Such a motion sensor 235 may employ an ultrawide band radio channel or a UHF (ultrahigh frequency) channel for wireless communication. Therefore, the radio transceiver system 150 may contain a UHF transceiver adapted to read out data from the motion sensor unit 235. Since radio signals emitted in the UHF frequency band typically have relatively long range, say in the order of 10ths of meters, it is advisable to somehow limit the range, for example via physical screening, a time-of-arri- val delimitation and / or signal strength.

[0049] Figure 3 shows a graph that represents a first milk-related parameter in the form of an amount of milk produced MP by the individual 130 as a function of time, which, in turn, here is expressed as the number of days in milk DIM for the individual 130. The amount of milk produced MP may further form a basis for the above-mentioned lactation-cycle-related parameter L. Figure 3 also shows a graph of the BCS for the individual 130 as a function of time, in terms of days in milk DIM.

[0050] As can be seen in Figure 3, the first and second graphs MP and BCS are correlated to one another in such a way that as the amount of milk produced MP increases, the BCS decreases. This is a typical characteristic for dairy animals. Namely, as the lactation cycle progresses, the amount of milk produced MP normally increases until it levels out on a plateau. During the same period, the BCS decreases due to the relatively large quantities of fat and energy required to enable the individual 130 to produce the milk. Subsequent to the plateau, the amount of milk produced MP normally decreases gradually, and this decrease continues until the individual 130 is set dry, typically for around two months, whereafter the lactation cycle restarted by a new pregnancy. Since there is a comparatively strong interrelationship between the amount of milk produced MP and the BCS as functions of time, the amount of milk produced MP may be used to set one or more criteria for whether the reference data record for the individual 130 shall be updated. Namely, any variations in the BCS are prone to show up as visible alterations in the individual’s appearance, especially its back portion. Therefore, by studying the amount of milk produced MP by the individual 130 it may be concluded if it is relevant to update the reference data record for the individual 130.

[0051] In the example illustrated in Figure 3, the BCS alters from a negative trend to a positive trend at a point in time t2. Between a point in time ti before t2 and a point in time ts after t2, the BCS is relatively stable, i.e. shows only minor variation. Consequently, it makes little sense to update the reference data record between time ti and ts. More precisely, the point in time t2, when the BCS graph shows a very small inclination iscs may be derived based on the MP graph because prior to t2, a positive inclination, or steepness, of the MP graph falls below a first threshold value IMPI . Analogously, the point in ts may be derived based on the MP graph by detecting when an inclination, or steepness, of the MP graph falls below a second threshold value iMP2. Referring to the example of Figure 3, a first updating criterion may be regarded as fulfilled for the individual 130 until the positive inclination of the MP graph falls below the first threshold value SMPI , and a second updating criterion may be regarded as fulfilled for the individual 130 after that the negative inclination of the MP graph has fallen below the second threshold value iMP2. In other words, the reference data record for the individual 130 may be updated before ti and after ts.

[0052] Figure 4 shows a graph that represents a second milk-related parameter in the form of a fat-compensated amount of milk produced FCMP for the individual 130 as a function of time, which also here is expressed as the number of days in milk DIM for the individual 130. As in Figure 3, Figure 4 further shows a graph of the BCS for the individual 130 as a function of time in terms of days in milk DIM. The fat-compensated amount of milk produced FCMP is correlated with the BCS in a manner similar to the amount of milk produced MP. Therefore, a period around the point in time t2, where the BCS is relatively stable, may be detected by detecting a point in time t4 when a positive inclination of the fat-compensated amount of milk produced FCMP falls below a third threshold value iFCMPi and a point in time t4 when a negative inclination of the fat-compensated amount of milk produced FCMP falls below a fourth threshold value ipcMP2. In further analogy to the example of Figure 3, a third updating criterion may here be regarded as fulfilled for the individual 130 until the positive inclination of the MP graph falls below the third threshold value IFMPI , and a fourth updating criterion may be regarded as fulfilled for the individual 130 after that the negative inclination of the MP graph has fallen below the fourth threshold value ipcMP2. In other words, the reference data record for the individual 130 may be updated before t4 and after ts.

[0053] Figure 5 shows a graph that represents a third milk-related parameter in the form of a duration MD of a milking session of the individual 130, which again is expressed as the number of days in milk DIM for the individual 130. Similar to Figures 3 and 4, Figure 5 shows a graph of the BCS for the individual 130 as a function of time, in terms of days in milk DIM. In contrast to the amount of milk produced MP and the fat-compensated amount of milk produced FCMP, the duration MD of the milking sessions typically do not vary much in the beginning of the lactation cycle. However, after a point in time te following the point in time t2, the duration MD of the milking sessions normally decrease as the milk production is lowered. The point in time te may be detected by checking if the negative inclination of the graph that expresses the duration MD of the milking sessions as a function of time falls below a fifth threshold value IMD. Analogous to the above, this may be regarded as fulfillment of one of the at least one updating criterion, meaning that after passing the point in time te, the reference data record for the individual 130 may be updated.

[0054] Figure 6 shows a graph that represents a fourth milk-related parameter in the form of a milk flow rate MF during a milking session for said one individual 130, which is expressed as the number of days in milk DIM for the individual 130. Analogous to Figures 3 to 5, Figure 6 shows a graph of the BCS for the individual 130 as a function of time in terms of days in milk DIM. As can be seen, the milk flow rate MF shows an interrelationship with the BCS similar to that of the duration MD of the milking sessions. Thus, an updating criterion may be defined based on a point in time t? when the negative inclination of a graph expressing the milk flow rate MF as a function of time falls below a sixth threshold value SMF, which here is assumed to occur at a point in time t? after t2. Consequently, one of the at least one updating criterion with respect to the individual 130 may be considered to be fulfilled after that the point in time t? has been passed.

[0055] Although the graphs illustrated in Figures 3 to 6 may be useful to determine whether one or more updating criteria are fulfilled, according to one embodiment of the invention, an updating criterion of the at least one updating criteria is simply tested against the number of days in milk for the individual 130. This is possible, since by experience, dairy animals are known to show relatively low variation in BCS / back-portion visual appearance during a period in the middle of the lactation cycle, say from around 80 DIM to around 150 DIM. During this period it is not very useful to update the reference data record for the individual, whereas updating preferably should be made both before and after said period. Hence, the updating criterion may for example be regarded as fulfilled prior to 80 DIM and after 150 DIM.

[0056] As an another alternative to the graphs illustrated in Figures 3 to 6, according to one embodiment of the invention, the BCS itself forms a basis for determining whether the at least one updating criterion is fulfilled. Here, the above-mentioned physiology parameter P reflects the BCS of the individual 130. Information reflecting the BCS is preferably received in the controller 110 via the data interface 180, for example from a processing unit of a milking installation associated with the system 100. Analogous to the above, the at least one updating criterion may be considered to be fulfilled if a slope iscs of a graph fulfills a steepness criterion, which second graph describes the BCS as a function of time, for example expressed in terms of days in milk DIM, i.e. in relation to a reference point-in-time for individual 130. Specifically, the at least one updating criterion may be considered to be fulfilled if a negative slope iscs of said graph is below a seventh threshold value, or a positive slope iscs of said graph exceeds an eighth threshold value.

[0057] According to one embodiment of the invention, the above-mentioned behavioral parameter B reflects an activity level of the individual 130, and at least one of the at least one updating criterion is considered to be fulfilled if an average value of the activity level of the individual 130 exceeds a threshold level during a predefined measurement period prior to a point in time when the 3D image data Dimg was registered. The activity level, which may be registered by the motion sensor 235 can be seen as an indicator of that the individual is in heat, i.e. that the animal is within the part of its oestrus cycle when it may become pregnant. This, in turn, is typically a phase when the visual appearance of the animal’s back- portion varies relatively fast. Therefore, at least one of the at least one updating criterion may be considered to be fulfilled if said activity-related threshold is exceeded, so that the reference data record for the individual 130 is allowed to be updated.

[0058] Returning now to Figure 1 , it is generally advantageous if the controller 1 10 is configured to effect the above procedure in an automatic manner by executing a computer program. Therefore, the controller 110 may include at least one processing unit 101 and a memory unit 105, i.e. non-volatile data carrier, storing a computer program 103, which, in turn, contains software for making the at least one processing unit 101 execute the actions mentioned in this disclosure when the computer program 103 is run on the at least processing unit 101.

[0059] In order to sum up, and with reference to the flow diagram in Figure 7, we will now describe the computer-implemented method according to the invention for identifying animals, which method is performed in the at least one processor 101 of the controller 100.

[0060] In a first step 710, it is checked if 3D image data has been obtained that represent a back portion of a preselected individual in a herd of animals. If so, a step 720 follows; and otherwise, the procedure loops back and stays in step 710.

[0061] In step 720, the 3D image data is processed to extract data describing an anatomy and / or a topology of the back portion of the preselected individual. Thereafter, in a step 730, the extracted data is matched against reference data records in a database, where each reference data record is associated with the identity of a particular one of the individuals in the herd of animals.

[0062] In a subsequent step 740, it is checked if a matching data record was found in the database. If so, a step 750 follows; and otherwise, the procedure loops back to step 710. In step 750, an identity is assigned to the preselected individual, which assigned identity is the identity that is associated with the matching reference data record in the database.

[0063] Thereafter, in a step 760, it is checked if at least one updating criterion is fulfilled, which at least one updating criterion relates to: a physiology parameter, a behavioral parameter, a reproductive-cycle parameter and / or a lactation-cycle-related parameter for the individual. If at least one of the at least one updating criterion is fulfilled, a step 770 follows; and otherwise, the procedure loops back to step 710. In step 770, it is checked if an identification of the individual has been obtained, which identification is independent from the 3D image data. For example an independent identification may be obtained from an RFID tag, an optically readable tag code or a motion sensor carried by the individual. If the independent identification was obtained, a step 780 follows; and otherwise, the procedure loops back to step 710.

[0064] In step 780, the reference data record for the individual is updated based on the 3D image data obtained in step 710. Thereafter, the procedure loops back to step 710.

[0065] The process steps described with reference to Figure 7 may be controlled by means of a programmed processor. Moreover, although the embodiments of the invention described above with reference to the drawings comprise processor and processes performed in at least one processor, the invention thus also extends to computer programs, particularly computer programs on or in a carrier, adapted for putting the invention into practice. The program may be in the form of source code, object code, a code intermediate source and object code such as in partially compiled form, or in any other form suitable for use in the implementation of the process according to the invention. The program may either be a part of an operating system, or be a separate application. The carrier may be any entity or device capable of carrying the program. For example, the carrier may comprise a storage medium, such as a Flash memory, a ROM (Read Only Memory), for example a DVD (Digital Video / Versatile Disk), a CD (Compact Disc) or a semiconductor ROM, an EPROM (Erasable Programmable Read-Only Memory), an EEPROM (Electrically Erasable Programmable Read-Only Memory), or a magnetic recording medium, for example a floppy disc or hard disc. Further, the carrier may be a transmissible carrier such as an electrical or optical signal which may be conveyed via electrical or optical cable or by radio or by other means. When the program is embodied in a signal, which may be conveyed, directly by a cable or other device or means, the carrier may be constituted by such cable or device or means. Alternatively, the carrier may be an integrated circuit in which the program is embedded, the integrated circuit being adapted for performing, or for use in the performance of, the relevant processes.

[0066] Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims.

[0067] The term “comprises / comprising” when used in this specification is taken to specify the presence of stated features, integers, steps or components. The term does not preclude the presence or addition of one or more additional elements, features, integers, steps or components or groups thereof. The indefinite article "a" or "an" does not exclude a plurality. In the claims, the word “or” is not to be interpreted as an exclusive or (sometimes referred to as “XOR”). On the contrary, expressions such as “A or B” covers all the cases “A and not B”, “B and not A” and “A and B”, unless otherwise indicated. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.

[0068] It is also to be noted that features from the various embodiments described herein may freely be combined, unless it is explicitly stated that such a combination would be unsuitable.

[0069] The invention is not restricted to the described embodiments in the figures, but may be varied freely within the scope of the claims.

Claims

Claims1 . A system (100) for identifying individuals in a herd of animals in which each individual has a known identity, the system (100) comprising: a first camera (140) configured to obtain three-dimensional, 3D, image data (Dimg) representing a back portion of a preselected individual (130) in the herd of animals, a database (160) comprising reference data records, wherein each reference data record is associated with the identity of a particular one of the individuals in the herd of animals, a controller (110) configured to: process the 3D image data (Dimg) to extract data (Dext) describing at least one of an anatomy and a topology of the back portion of the preselected individual (130), match the extracted data (Dext) against the reference data records in a database (160), assign an identity to the preselected individual (130) in response to a match between the extracted data (Dext) and one of the reference data records in the database (160), which assigned identity is the identity that is associated with the matching reference data record in the database (160), and update, repeatedly, each of the reference data records in the database (160), which updating comprises obtaining 3D image data (Dimg) representing the back portion of one of the individuals in the herd of animals together with an identification (ID) of said one individual (130), which identification (ID) is independent from the 3D image data (Dimg), characterized in that the controller (110) is configured to update the reference data record in the database (160) if and only if at least one updating criterion is fulfilled for said one individual (130), which at least one updating criterion relates to at least one of: a physiology parameter (P), a behavioral parameter (B), a reproductive-cycle parameter (R) and a lactation-cycle-related parameter (L) for said one individual (130).

2. The system according to claim 1 , wherein the controller (110) is configured to derive the lactation-cycle-related parameter (L) based on at least one milk-related parameter (MP, FCMP, MD, MF) for said one individual (130), which at least one milk-related parameter reflects at least one of: an amount of milk produced (MP) by said one individual (130), a fat-compensated amount of milk produced (FCMP) for said one individual (130), a duration (MD) of a milking session of said one individual (130), a milk flow rate (MF) during a milking session for said one individual (130), and a number of days in milk for said one individual (130).

3. The system according to claim 2, wherein at least one of the at least one updating criterion is considered to be fulfilled if a slope (IBCS) of at least one first graph fulfills at least one steepness criterion (SMPI , iMP2j IFCMPI , ipcMP2; IMD ;, which at least one first graph describes the amount of milk produced (MP) by said one individual (130), the fat-compensated amount of milk produced (FCMP) for said one individual (130), the duration (MD) of a milking of said one individual (130), and / or the milk flow rate (MF) during a milking session for said one individual (130) respectively as a function of time (DIM) in relation to a reference point-in-time (0) for said one individual (130).

4. The system according to any one of the preceding claims, wherein the behavioral parameter (B) reflects an activity level of said one individual (130); and at least one of the at least one updating criterion is considered to be fulfilled if an average value of the activity level of said one individual (130) exceeds a threshold level during a predefined measurement period prior to a point in time when the 3D image data (Dimg) was registered.

5. The system according to any one of the preceding claims, wherein the physiology parameter (P) reflects a body conditionscore (BCS) of said one individual (130), and at least one of the at least one updating criterion is considered to be fulfilled if a slope (IBCS) of a second graph fulfills a steepness criterion, which second graph describes the body condition score (BCS) as a function of time (DIM) in relation to a reference point-in-time (0) for said one individual (130).

6. The system according to any one of the preceding claims, wherein the extracted data (Dext) comprises at least one of: a set of image features derived from the 3D image data (Dimg) , which set of image features characterize the anatomy and / or the topology of the back portion of the preselected individual (130), and a compressed version of the 3D image data (Dimg) .

7. The system according to any one of the preceding claims, comprising a radio transceiver system (150) configured to read out a code from an RFID tag (135) that is carried by said one individual (130), which code forms a basis for the identification (ID) that is independent from the 3D image data (Dimg).

8. The system according to any one of the preceding claims, comprising an imaging subsystem (170) configured to register image data comprising an optically readable code on a tag (245) that is carried by said one individual (130), which optically readable code forms a basis for the identification (ID) that is independent from the 3D image data (Dimg) .

9. The system according to any one of the preceding claims, comprising at least one data interface (180) configured to obtain at least one of the physiology parameter (P), the behavioral parameter (B), the reproductive-cycle parameter (R) and the lac- tation-cycle-related parameter (L) for said one individual (130).

10. A computer-implemented method for identifying individuals in a herd of animals in which each individual has a known identity,which method is executed in at least one processing unit (101 ) of a controller (110) and which method comprises: obtaining 3D image data (Dimg) representing a back portion of a preselected individual (130) in the herd of animals, processing the 3D image data (Dimg) to extract data (Dext) describing at least one of an anatomy and a topology of the back portion of the preselected individual (130), matching the extracted data (Dext) against reference data records in a database (160), wherein each reference data record is associated with the identity of a particular one of the individuals in the herd of animals, assigning an identity to the preselected individual (130) in response to a match between the extracted data (Dext) and one of the reference data records in the database (160), which assigned identity is the identity that is associated with the matching reference data record in the database (160), and updating, repeatedly, each of the reference data records in the database (160), which updating comprises obtaining 3D image data (Dimg) representing the back portion of one of the individuals in the herd of animals together with an identification (ID) of said one individual (130), which identification (ID) is independent from the 3D image data (Dimg) , characterized by updating the reference data record in the database (160) if and only if at least one updating criterion is fulfilled for said one individual (130), which at least one updating criterion relates to at least one of: a physiology parameter (P), a behavioral parameter (B), a reproductive-cycle parameter (R) and a lactation- cycle-related parameter (L) for said one individual (130).

11. The method according to claim 10, comprising: deriving the lactation-cycle-related parameter (L) based on at least one milk-related parameter (MP, FCMP, MD, MF) for said one individual (130), which at least one milk-related parameter reflects at least one of: an amount of milk produced (MP) by said one individual (130), a fat-compensated amount of milk produced (FCMP) for said one individual (130), a duration (MD) of a milkingof said one individual (130), and a milk flow rate (MF) for said one individual (130).

12. The method according to claim 11 , wherein at least one of the at least one updating criterion is considered to be fulfilled if a slope (IBCS) of at least one first graph fulfills at least one steepness criterion (SMPI , iMP2j IFCMPI , ipcMP2; IMD ;, which at least one first graph describes the amount of milk produced (MP) by said one individual (130), the fat-compensated amount of milk produced (FCMP) for said one individual (130), the duration (MD) of a milking of said one individual (130), and / or the milk flow rate (MF) for said one individual (130) respectively as a function of time (DIM) in relation to a reference point-in-time (0) for said one individual (130).

13. The method according to any one of claims 10 to 12, wherein the behavioral parameter (B) reflects an activity level of said one individual (130); and at least one of the at least one updating criterion is considered to be fulfilled if an average value of the activity level of said one individual (130) exceeds a threshold level during a predefined measurement period prior to a point in time when the 3D image data (Dimg) was registered.

14. The method according to any one of claims 10 to 13, wherein the physiology parameter (P) reflects a body condition score (BCS) of said one individual (130), and at least one of the at least one updating criterion is considered to be fulfilled if a slope (IBCS) of a second graph fulfills a steepness criterion, which second graph describes the body condition score (BCS) as a function of time (DIM) in relation to a reference point-in-time (0) for said one individual (130).

15. The method according to any one of claims 10 to 14, wherein the identification (ID) being independent from the 3D image data (Dimg) comprises obtaining a code from a tag (135, 245) that is carried by said one individual (130), which code is obtained overa wireless interface (155, 175).

16. The method according to claim 15, wherein the obtaining of the code via the wireless interface (155, 175) comprises at least one of a radio-based read-out from said tag (135) and an optical reading of said tag (135, 245).

17. A computer program (103) loadable into a non-volatile data carrier (105) communicatively connected to a processing unit (101 ), the computer program (103) comprising software for executing the method according to any of claims 10 to 16 when the computer program (103) is run on the processing unit (101 ).

18. A non-volatile data carrier (105) containing the computer program (103) of the claim 17.

Images