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

Abstract

At a first stage, a respective identity of the individuals in a sub-group of animals is determined by matching a respective code (C[1]) obtained from a radio-based read-out of a respective tag carried by each individual against a set of reference codes (DR) in a source database (115). At a second stage, respective animal-related data (Dimg) in the form of 2D and / or 3D image data are registered for each individual in the subgroup. The registered ima-ge data (Dimg) are matched against a subset of reference data records (dr) in a temporary database (125) containing only reference data records with reference image data relating to the individuals in the subgroup of animals. A confirmed identity is assigned to one of the individuals in the subgroup in response to a match between the registered image data and the reference image data, which confirmed identity is an animal identity associated with the matching reference data record.

Description

[0001] CONTROLLER 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 controller according to the preamble of claim 1 , a corresponding computer- implemented method and an animal identification system including said controller. 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 also 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. A few prior-art examples of automatic identification systems are listed below.

[0006] EP 3 446 563 discloses a system wherein a transmission signal is transmitted from a signal transmission device attached to a domestic animal and relayed to the information processing server by at least one or more signal relay devices installed to correspond to a rearing region of the domestic animal. The system includes a transmission device identifier for specifying the signal transmission device and a relay device identifier for specifying the signal relay device. The information processing server includes: a control unit configured to determine whether or not the domestic animal stays in a rearing region corresponding to a rearing stage on a basis of the relay device identifier and individual infor- mation regarding the domestic animal corresponding to the transmission device identifier; and a notification information generation unit configured to generate notification information for controlling notification in an information terminal on a basis of the determination by the control unit.

[0007] EP 2 983 465 shows an animal monitoring system for determining feed consumption of one or more animals feeding at a feeding area. The system comprises an imaging unit for range imaging the feeding area, identification means configured to uniquely identify each feeding animal, and processing means configured for assessing the amount of feed consumed by each identified animal by determining the reduction of feed in subsequent images of the feeding area in front of each identified animal.

[0008] WO 2009 / 016183 reveals a device and method for providing information about animals walking through an animal passage. The information comprises the number of animals walking through the animal passage. For capturing animal data about animals walking through the animal passage, a detection device is used that has a sensor device connected to a processor. Further, an analysis device is used for recognizing animals in the data / signals captured by the sensor device for the purpose of outputting counter impulses when animals are detected in said signals. The sensor device is designed for producing 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.

[0009] WO 2014 / 067897 describes a system and a method for real time detection of the position and behavior of a plurality of animals. The system comprises radio transmitter tags carried by the animals, where sensors receive signals from the tags. Based on time delay of received signals at the sensors from the tags, the system performs calculation the position of the animals. Real time positioning of animal in a limited area is achieved by letting the system comprise a computer-based real time position calculation of the position of the animals. The system comprises a plurality of fixed refe- rence radio transmitter tags. Based on received signal from the reference tags, The system continuously performs a calibration of the measured position of the animals. By using fixed reference radio transmitters together with tags that are performing equal radio transmission functions, it is possible, by fixed placed sensors, to receive first signals from the animal tags. By also using signals from the fixed reference radio transmitter tags, it further possible to perform an automatic calibration of all received signals from animals.

[0010] Thus, various kinds of strategies are known for identifying and counting animals, some of which strategies involve image-based methods. However, the data matching needed to attain a highly reliable identification typically demands processing of large amounts of data. The matching process therefore becomes either expensive or 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 automatic identification of animals.

[0013] According to one aspect of the invention, the object is achieved by a controller for identifying animals. The controller is configured to perform first and second stages of identification. At the first stage, a respective identity is determined of each individual in a subgroup of animals by matching a respective code obtained from a radio-based read-out of a respective tag carried by each individual in the subgroup of animals against a set of reference codes in a source database. The source database contains a respective reference code for each individual in a herd of animals in which the subgroup of animals is comprised. The second stage identification of each individual in the subgroup of animals involves matching respective registered animal-related data for each individual in the subgroup against a subset of reference data records selected from the source database. The subset of reference data re- cords is selected based on the first-stage identification such that the subset of reference data records only contains reference data records relating to the individuals in the subgroup of animals. The respective registered animal-related data for each individual in the subgroup contains registered two-dimensional, 2D, image data and / or registered three-dimensional, 3D, image data. The subset of reference data records contains reference image data for each individual, for example in the form of feature vectors, pre-selected image points and / or complete images. Each reference data record in the subset is associated with an animal identity. The controller is configured to perform the second-stage identification through a process that involves matching the registered image data against the reference image data, and assigning a confirmed identity to one of the individuals in the subgroup in response to a match between the registered image data and the reference image data, which confirmed identity is the animal identity associated with the matching reference data record.

[0014] This controller is advantageous because the selected subset of reference data records renders the matching process in the second stage highly efficient, especially in comparison to if such matching would have been performed with respect to all the data records in the source database. Namely, the source database may contain a very large number of data records, say in the order of many thousands, and typically, each image-based matching step is relatively processing intense.

[0015] According to one embodiment of this aspect of the invention, the registered image data represents a respective back part of each individual in the subgroup of animals and / or a respective head / face of each individual in the subgroup of animals. Consequently, the subset of reference data records contains reference image data representing a respective back part of each individual in a subgroup of animals and / or a respective head / face of each individual in a subgroup of animals respectively. Analogous to the above, the reference data records may contain reference image data in the form of feature vectors, pre-selected image points and / or complete images.

[0016] According to another embodiment of this aspect of the invention, the controller is configured to perform a repetition of the first-stage identification of the respective identities of each individual in the subgroup of animals, and based thereon update the subset of reference data records relating to the individuals in the subgroup of animals. Namely, in most cases, there is a gradual flow of animals through an entry area that temporarily holds the subgroup of animals. Therefore, the subset of reference data records needs to be updated repeatedly.

[0017] Preferably, the controller is configured to perform the repetition of the first-stage identification of the respective identities of each individual in the subgroup of animals in response to assigning the confirmed identity to one of the individuals in the subgroup because once an individual has been identified, the individual in question does not normally need to be included in the selected subset any longer.

[0018] Analogously, according to yet another embodiment of this aspect of the invention, after having assigned the confirmed identity to said one of the individuals in the subgroup, the controller is further configured to delete a reference data record from the subset of reference data records, which deleted reference data record relates to the individual to whom the confirmed identity was assigned. Consequently, it is ensured that the subset of reference data records exclusively holds those reference data records that may be useful to identify the individuals at the second stage.

[0019] According to another aspect of the invention, the object is achieved by an animal identification system that contains: a first wireless reader arrangement, a second wireless reader arrangement, a physically delimited area, a source database, a temporary database and a central unit. The first wireless reader arrangement is configured to, based on radio energy, read out a respective code from a respective tag carried by each individual in a herd of ani- mals, the radio energy having a range that covers an entry area adapted to temporarily hold a subgroup of animals comprising at least two individuals from the herd of animals at the same time. The physically delimited area is configured to hold one and only one individual from the subgroup of animals at the same time. The second wireless reader arrangement is configured to register respective animal-related data for each individual in the subgroup of animals while the individual is located in the physically delimited area. The source database includes reference data records which each comprises a respective reference code for each individual in the herd of animals. The central unit is configured to obtain the respective codes from the respective tags read out via the first wireless reader arrangement. The central unit is configured to perform a first-stage identification to determine a respective identity of each individual in the subgroup of animals by matching said codes against the set of reference codes in the source database. The central unit is further configured to select a subset of reference data records from the source database based on the first- stage identification of the respective identities of each individual in the subgroup such that the subset of reference data records only contains reference data records relating to the individuals in the subgroup of animals; and store the subset of reference data records in the temporary database. Additionally, the central unit is configured to perform a second-stage identification of each individual in the subgroup of animals. The second-stage identification here involves using the second wireless reader arrangement and matching the respective registered animal-related data for each individual in the subgroup of animals against the subset of reference data records in the temporary database. Specifically, according to the invention, the second wireless reader arrangement includes a camera configured to register animal-related data comprising registered 2D and / or 3D image data. The central unit includes a controller for identifying animals, which controller is configured to perform the second-stage identification of each individual in the subgroup of animals through a procedure that involves matching respective registered animal-related data for each individual in the subgroup against the subset of reference data records. The subset of reference data records contains reference image data for each individual in the subgroup and each reference data record is associated with an animal identity. In response to a match between the registered image data and the reference image data, the controller is configured to assign a confirmed identity, which is the animal identity associated with the matching reference data record.

[0020] The advantages of this animal identification system, as well as the preferred embodiments thereof, are apparent from the discussion above with reference to the proposed controller.

[0021] According to one embodiment of this aspect of the invention, the entry area is a bridge to a rotating platform and the physically delimited area is a bail that is associated with milking equipment configured to extract milk from a milk-producing animal. Thus, it may be ensured that each dairy animal is adequately identified in connection with milking thereof.

[0022] Alternatively, the physically delimited area is a general passageway or a sort gate, for example into a feeding area. Hence, the invention may be applied anywhere in a farm, where reliable animal identification is required.

[0023] In general, each dairy animal requires two square meters of space. For example, in a rotating milking platform, the bridge should accommodate at least two animals simultaneously. Therefore, according to one embodiment of the invention, the entry area is at least four square meters. A holding pen for dairy cows, however, may simultaneously accommodate as many as two hundred individuals or more. Consequently, according to another embodiment of the invention, the entry area is at least four hundred square meters.

[0024] Additionally, it is preferred that the camera in the second wireless reader arrangement is arranged such that the image data covers a predefined portion of the physically delimited area, since there- by, a most relevant position for identification may be conveniently automatically monitored.

[0025] According to yet another aspect of the invention, the object is achieved by a computer-implemented method, which is performed in a processing unit in a controller, which controller, in turn, is arranged to identify animals. The method involves performing a first-stage identification to determine a respective identity of each individual in a subgroup of animals by matching a respective code obtained from a radio-based read-out of a respective tag carried by each individual in the subgroup of animals against a set of reference codes in a source database containing a respective reference code for each individual in a herd of animals in which the subgroup of animals is included. The method further involves performing a second-stage identification of each individual in the subgroup of animals. The second-stage identification, in turn, involves matching respective registered animal-related data for each individual in the subgroup against a subset of reference data records. The subset of reference data records is selected from the source database based on the first-stage identification such that the subset of reference data records only contains reference data records relating to the individuals in the subgroup of animals. For each individual in the subgroup, the respective registered animal- related data contains registered 2D and / or 3D image data, Analogously, the subset of reference data records contains reference image data for each individual in the subgroup, and each reference data record is associated with an animal identity. Additionally, the method involves performing the second-stage identification through a process that involves matching the registered image data against the reference image data, and assigning a confirmed identity to one of the individuals in the subgroup in response to a match between the registered image data and the reference image data, which confirmed identity is the animal identity associated with the matching reference data record.

[0026] The advantages of this method, as well as the preferred embodiments thereof, are apparent from the discussion above with refe- rence to the proposed controller and system.

[0027] 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 a processing unit. The computer program includes software for executing the above method when the program is run on the processing unit.

[0028] According to another aspect of the invention, the object is achieved by a non-volatile data carrier containing the above computer program.

[0029] Further advantages, beneficial features and applications of the present invention will be apparent from the following description and the dependent claims.

[0030] BRIEF DESCRIPTION OF THE DRAWINGS

[0031] 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.

[0032] Figure 1 schematically illustrates an entry area and a physically delimited area respectively together with an animal identification system according to a first embodiment of the invention;

[0033] Figure 2 shows an animal carrying a tag according to one embodiment of the invention;

[0034] Figure 3 schematically illustrates an entry area and a physically delimited area respectively together with an animal identification system according to a second embodiment of the invention; and

[0035] Figure 4 illustrates, by means of a flow diagram, the general method for identifying animals according to the invention. DETAILED DESCRIPTION

[0036] Figure 1 shows an entry area 117, a physically delimited area 123 respectively and an animal identification system according to a first embodiment of the invention. Figure 2 shows an individual animal 230 carrying a tag 235 according to one embodiment of the invention.

[0037] The animal identification system includes a first wireless reader arrangement 110, a second wireless reader arrangement 120, a source database 1 15, a temporary database 125 and a central unit 1000.

[0038] The first wireless reader arrangement 110 is configured to, based on radio energy, read out a respective code C[i] from a respective tag 235 carried by each individual 230 in a herd of animals.

[0039] The first wireless reader arrangement 110 may for example be an RFID (radio frequency identification) system, 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 first wireless reader arrangement 110 may be 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 any case, according to the invention, the radio energy of the first wireless reader arrangement 110 has a range R that covers the entry area 117, which, in turn, is adapted to temporarily hold a subgroup of animals 130 including at least two individuals 230 from the herd of animals at the same time. If the entry area 117 is a holding pen for dairy cows, as exemplified in Figure 1 , the entry area 117 may simultaneously accommodate as many as two hundred individuals, or even more. In such a case, the entry area 117 may be at least four hundred square meters. The tag 235 is implemented in a technology suitable for the first wireless reader arrangement 110. This means that if the first wireless reader arrangement 110 is an RFID system, the tag 235 may be an RFID transponder. Analogously, if the first wireless reader arrangement 110 is a UWB communication system, the tag 235 contains a UWB transceiver.

[0040] Regardless of how large the entry area 117 is the physically delimited area 123 is configured to hold one and only one individual 230 from the subgroup of animals 130 at the same time. For example, the physically delimited area 123 may be a passageway as shown in Figure 1 , or a sort gate.

[0041] A second wireless reader arrangement 120 is configured to register respective animal-related data Dimgfor each individual 230 in the subgroup of animals 130 while the individual 230 is located in the physically delimited area 123. The second wireless reader arrangement 120 includes a camera, e.g. a TOF (time-of-flight) camera configured to register animal-related data Dimg that contains registered 3D image data of the individual 230. Alternatively, or additionally, the camera may register 2D image data of the individual 230 in the physically delimited area 123.

[0042] The source database 115 contains reference data records which each includes a respective reference code DR for each individual in the herd of animals. Each reference data record also includes reference image data for each individual 230 in the herd along with an associated animal identity. Here, the reference image data may for example be represented by feature vectors, pre-selected image points and / or complete images of each individual 230.

[0043] The central unit 1000 includes a controller 100, which is configured to identify animals according to the below.

[0044] The controller 100 is configured to obtain the respective codes C[i] from the respective tags 235 of the individuals in the entry area 117. The codes C[i] are read out via the first wireless reader arrangement 110 and each code C[i] is uniquely linked to a parti- cular individual 230 in the herd, preferably via its respective reference data record in the source database 115.

[0045] The controller 100 is configured to perform a first-stage identification to determine a respective identity IDi , ID2, ID3, ID4 and IDs respectively of each individual in the subgroup of animals 130 by matching the codes C[i] against the set of reference codes DR in the source database 115. Each code C[i] is preferably a simple symbol string. This namely renders the matching process relatively uncomplicated and non-demanding in terms of the processing resources required.

[0046] The controller 100 is also configured to select a subset of reference data records dr from the source database 1 15 based on the first-stage identification of the respective identities ID1, ID2, ID3, ID4 and IDs of each individual 230 in the subgroup 130. Here, the selection is made such that the subset of reference data records dr only contains reference data records relating to the individuals 230 in the subgroup of animals 130. As mentioned above, the subgroup 130 may for example contain 2 - 200 individuals 230. In any case, the individuals 230 in the subgroup 130 typically constitutes only a fraction of the total number of animals in the entire herd.

[0047] Moreover, the controller 100 is configured to store the subset of reference data records dr in the temporary database 125. Thereafter, the controller 100 is configured to perform a second-stage identification of each individual 230 in the subgroup of animals 130 through a procedure that involves matching respective registered animal-related data Dimgfor each individual 230 in the subgroup 130 against the subset of reference data records dr.

[0048] Since the source database 115 contains reference data records with reference image data, e.g. represented by feature vectors, pre-selected image points and / or complete images of each individual 230 which reference data record each is associated with a respective animal identity, the subset of reference data records dr likewise contain reference image data, e.g. represented by feature vectors, pre-selected image points and / or complete images of each individual 230 in the subgroup 130, which each reference data record is associated with a respective animal identity.

[0049] In response to a match between the registered image data and the reference image data, the controller 100 is configured to assign a confirmed identity to the one of the individuals in the subgroup 130 whose registered image data Dimgis associated with the matching reference data record.

[0050] Depending on where the camera of the second wireless reader arrangement 120 is positioned and the view angle of the camera, according to one embodiment of the invention, the registered image data may represent a respective back part of each individual in the subgroup of animals 130 and / or a respective head / face of each individual in the subgroup of animals 130. Analogously, the subset of reference data records dr may contain reference image data representing a respective back part of each individual in a subgroup of animals 130 and / or a respective head / face of each individual in a subgroup of animals 130 respectively.

[0051] In most cases, the subgroup of animals 130 gradually alters over time with respect to its members because different individuals 230 from the herd pass via the entry area 117, through the physically delimited area 123 and further in to a particular part of a farm environment.

[0052] Therefore, according to one embodiment of the invention, the controller 100 is configured to perform a repetition of the first-stage identification of the respective identities IDi, ID2, ID3, ID4 and IDs of each individual in the subgroup of animals 130. Based thereon, in turn, the controller 100 is configured to update the subset of reference data records dr relating to the individuals 230 in the subgroup of animals 130. Typically, this means adding data records dr relating to one or more individuals 230 to the temporary database 125.

[0053] For instance, the controller 100 may be configured to perform the repetition of the first-stage identification of the respective identities IDi , ID2, ID3, ID4 and IDs of each individual in the subgroup of animals 130 in response to assigning the confirmed identity to one of the individuals 230 in the subgroup 130. Namely, at this point in time, it is reasonable to expect that another individual from the subgroup 130 will pass from the entry area 117 in to the physically delimited area 123. Therefore, it is appropriate to update the temporary database 125.

[0054] Further, for similar reasons, after having assigned the confirmed identity to said one of the individuals in the subgroup 130 it is reasonable to reduce the temporary database 125. Specifically, according to one embodiment of the invention, after having assigned the confirmed identity to one of the individuals in the subgroup 130, the controller 100 is configured to delete a reference data record from the subset of reference data records dr, which deleted reference data record relates to the individual 230 to whom the confirmed identity was assigned. As a result, the temporary database 125 is kept up to date with respect to the current individuals in the subgroup 130. This is beneficial considering the processing resources required to perform the matching of the respective registered animal-related data Dimgfor each individual 230 in the subgroup 130 against the subset of reference data records dr, since the contents of the temporary database 125 is limited to those individuals that are currently located in the entry area 117.

[0055] It is generally advantageous if the controller 100 is configured to effect the above-described procedure in an automatic manner by executing a computer program. Therefore, the controller 100 may include a memory unit 105, i.e. non-volatile data carrier, storing a computer program 103, which, in turn, contains software for making processing circuitry in the form of at least one processor 101 in the controller 100 execute the actions mentioned in this disclosure when the computer program 103 is run on the at least one processor 101 .

[0056] Figure 3 schematically illustrates an entry area 317 and a physi- cally delimited area 323 respectively according to a second embodiment of the invention. The animal identification system, however, is the same as described above referring to Figures 1 and 2. In other words, the reference symbols in Figure 3 that also occur in Figure 1 designate the same entities, signals and parameters respectively as described above in relation to Figure 1.

[0057] In Figure 3, the animal identification system is adapted to identify dairy animals in connection with entering a rotating milking platform 300. Here, therefore, the entry area 317 is a bridge to the rotating platform 300 and the physically delimited area 323 is a bail that is associated with milking equipment MP1 configured to extract milk from a milk-producing animal.

[0058] As mentioned earlier, the bridge is relatively small in comparison to the holding pen shown in Figure 1 . For example, the entry area 317 in Figure 3 may be adapted to simultaneously accommodate two to four animals and may thus be around 4 - 8 m2in size.

[0059] In addition to the milking equipment MP1 , Figure 3 shows milking equipment MP2, MP3, MP4 and MP5 respectively that will be occupied after that the milking equipment MP1 which is represented by the physically delimited area 323 has been occupied by an individual 230 to whom the controller 100 has hopefully been able to assign a confirmed identity. The rotating platform 300 is presumed to move in a counter clockwise direction F, which, here, means that after that a particular bail, say the one associated with the milking equipment MP1 , has passed behind a so-called kick rail 321 that encloses the rotating platform 300, an animal located in said particular bail cannot reenter the entry area 317. At a somewhat later point in time, an animal located in the bail associated with the milking equipment MP2 cannot return to the entry area 317, and so on. Thus, animals gradually progress from the bridge represented by the entry area 317 onto the rotating platform 300, where each of the individuals 230 is assigned a confirmed identity in the physically delimited area 323. Since, once an individual has entered the physically delimited area 323, the individual 230 can- not divert therefrom, it is advantageous to arrange the second wireless reader arrangement 120 to register the image data Dimgof the individual 230 to be identified when the individual 230 is located in the physically delimited area 323.

[0060] Preferably, the camera included in the second wireless reader arrangement 120 is arranged, i.e. has such location and field of view that the image data Dimg registered by the camera covers a predefined portion of the physically delimited area 123 or 323 respectively in which predefined portion each animal is expected to be located when the animal in question cannot back out from the physically delimited area 123 / 323 and where the registered image data Dimg is expected to provide a fair representation of the individual 230.

[0061] In order to sum up, and with reference to the flow diagram in Figure 4, 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.

[0062] In a first step 410, a first-stage identification is performed in which a respective identity is determined of each individual in a subgroup of animals from the herd of animals. Here, the subgroup of animals is presumed to include at least two individuals that are temporarily located in an entry area, where a respective tag carried by each of said individuals is located within a range of the radio energy of a first wireless reader arrangement. The first-stage identification involves matching a respective code obtained from a radio-based read-out of the respective tag carried by the individuals in the subgroup of animals against a set of reference codes in a source database containing a respective reference code for each individual in a herd of animals in which the subgroup of animals is comprised.

[0063] In a subsequent step 420, a subset of reference data records is selected from the source database, which, in addition to the refe- rence codes also contains reference image data and an associated animal identity of each individual in the herd. The subset of reference data records is selected based on the first-stage identification, such that the subset of reference data records only contains reference data records relating to the individuals in the subgroup of animals identified in step 410.

[0064] Thereafter, in a step 430, the subset of reference data records is stored in a temporary database. Then, in a step 440, it is checked if an individual from the subgroup of animals is located in a physically delimited area that is configured to hold one and only one individual and from which physically delimited area the individual cannot return to the entry area. If it is found that an individual is located in the physically delimited area, a step 450 follows. Otherwise, the procedure loops back and stays in step 440.

[0065] In step 450, a second-stage identification is performed in which an attempt is made to identify the individual that is located in the physically delimited area based on registered 2D and / or 3D image data that represent said individual. Thus, the second-stage identification involves matching the registered image data against the subset of reference data records in the temporary database.

[0066] A following step 460 checks if a match can be found between the registered image data and a particular reference data record in the temporary database. If such a match is found, a step 470 follows. Otherwise, the procedure loops back to step 410.

[0067] In step 470, a confirmed identity is assigned to the animal that is associated with the matching reference data record. Thereafter, the procedure loops back to step 410.

[0068] The process steps described with reference to Figure 4 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.

[0069] 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.

[0070] 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.

[0071] 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.

[0072] 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 controller (100) for identifying animals, which controller (100) is configured to: perform a first-stage identification to determine a respective identity (IDi , ID2, ID3, ID4, IDs) of each individual in a subgroup of animals (130) by matching a respective code (C[1 ]) obtained from a radio-based read-out of a respective tag (235) carried by each individual (230) in the subgroup of animals (130) against a set of reference codes (DR) in a source database (115) comprising a respective reference code for each individual (230) in a herd of animals in which the subgroup of animals (130) is comprised, and perform a second-stage identification of each individual (230) in the subgroup of animals (130), which second-stage identification involves matching respective registered animal-related data (Dimg) for each individual (230) in the subgroup (130) against a subset of reference data records (dr), which subset of reference data records (dr) is selected from the source database (115) based on the first-stage identification such that the subset of reference data records (dr) only contains reference data records relating to the individuals (230) in the subgroup of animals (130), characterized in that the respective registered animal-related data (Dimg) for each individual (230) in the subgroup (130) comprises at least one of registered two-dimensional, 2D, image data and registered three- dimensional, 3D, image data, the subset of reference data records (dr) comprises reference image data for each individual (230) in the subgroup (130), wherein each reference data record is associated with an animal identity, and the controller (100) is configured to perform the second-stage identification through a process that involves matching the registered image data against the reference image data, and assigning a confirmed identity to one of the individuals in the subgroup (130) in response to a match between the registered image data and the reference image data, which confirmed identity is the animal identity associated with the matching reference data record.

2. The controller (100) according to claim 1 , wherein: the registered image data represents at least one of a respective back part of each individual in the subgroup of animals (130) and a respective head / face of each individual in the subgroup of animals (130), and the subset of reference data records (dr) comprises reference image data representing at least one of a respective back part of each individual in a subgroup of animals (130) and a respective head / face of each individual in a subgroup of animals (130) respectively.

3. The controller (100) according to any of claims 1 or 2, wherein the controller (100) is configured to: perform a repetition of the first-stage identification of the respective identities (IDi , ID2, ID3, ID4, IDs) of each individual in the subgroup of animals (130), and based thereon update the subset of reference data records (dr) relating to the individuals (230) in the subgroup of animals (130).

4. The controller (100) according to claim 3, wherein the controller (100) is configured to perform the repetition of the first-stage identification of the respective identities (ID1, ID2, ID3, ID4, IDs) of each individual in the subgroup of animals (130) in response to assigning the confirmed identity to one of the individuals (230) in the subgroup (130).

5. The controller (100) according to claim 4, wherein, after having assigned the confirmed identity to said one of the individuals in the subgroup (130), the controller (100) is further configured to delete a reference data record from the subset of reference data records (dr), which deleted reference data record relates to the individual (230) to whom the confirmed identity was assigned.

6. An animal identification system comprising: a first wireless reader arrangement (110) configured to, based on radio energy, read out a respective code (C[i]) from a res-pective tag (235) carried by each individual (230) in a herd of animals, the radio energy having a range (R) that covers an entry area (117, 317) adapted to temporarily hold a subgroup of animals (130) comprising at least two individuals (230) from the herd of animals at the same time, a physically delimited area (123, 323) configured to hold one and only one individual (230) from the subgroup of animals (130) at the same time, a second wireless reader arrangement (120) configured to register respective animal-related data (Dimg) for each individual (230) in the subgroup of animals (130) while the individual (230) is located in the physically delimited area (123, 323), a source database (115) with reference data records which each comprises a respective reference code (DR) for each individual in the herd of animals, a temporary database (125), and a central unit (1000) configured to: obtain said respective codes (C[i]) from the respective tags (235) read out via the first wireless reader arrangement (110), and perform a first-stage identification to determine a respective identity (IDi, ID2, ID3, ID4, IDs) of each individual in the subgroup of animals (130) by matching said codes (C[i]) against the set of reference codes (DR) in the source database (115), select a subset of reference data records (dr) from the source database (1 15) based on the first-stage identification of the respective identities (ID1 , ID2, ID3, ID4, IDs) of each individual (230) in the subgroup (130) such that the subset of reference data records (dr) only contains reference data records relating to the individuals (230) in the subgroup of animals (130), store the subset of reference data records (dr) in the temporary database (125), and perform a second-stage identification of each individual (230) in the subgroup of animals (130), which second- stage identification involves using the second wirelessreader arrangement (120) and matching the respective registered animal-related data (Dimg) for each individual (230) in the subgroup of animals (130) against the subset of reference data records (dr) in the temporary database (125), characterized in that the second wireless reader arrangement (120) comprises a camera configured to register animal-related data (Dimg) comprising at least one of registered two-dimensional, 2D, image data and registered three-dimensional, 3D, image data, and the central unit (1000) comprises a controller (100) for identifying animals, which controller (100) is configured to: perform the second-stage identification of each individual (230) in the subgroup of animals (130) through a procedure that involves matching respective registered animal-related data (Dimg) for each individual (230) in the subgroup (130) against the subset of reference data records (dr), the subset of reference data records (dr) comprising reference image data for each individual (230) in the subgroup (130), wherein each reference data record is associated with an animal identity, and assigning a confirmed identity to one of the individuals in the subgroup (130) in response to a match between the registered image data and the reference image data, which confirmed identity is the animal identity associated with the matching reference data record.

7. The animal identification system according to claim 6, wherein the entry area (317) is a bridge to a rotating platform (300) and the physically delimited area (323) is a bail that is associated with milking equipment (MP1 ) configured to extract milk from a milk-producing animal.

8. The animal identification system according to claim 6, wherein the physically delimited area (123) is a passageway or a sort gate.

9. The animal identification system according to any of claims 6 to 8, wherein the camera comprised in the second wirelessreader arrangement (120) is arranged such that the image data (Dimg) covers a predefined portion of the physically delimited area (123, 323).

10. The animal identification system according to any of claims 6 to 9, wherein the entry area (317) is at least four square meters.

11. The animal identification system according to any of claims 6 to 10, wherein the entry area (117) is at least four hundred square meters.

12. A computer-implemented method for identifying animals, which method is executed in a processing unit (101 ) in a controller (100), the method comprising: performing a first-stage identification to determine a respective identity (IDi , ID2, ID3, ID4, IDs) of each individual in a subgroup of animals (130) by matching a respective code (C[1 ]) obtained from a radio-based read-out of a respective tag (235) carried by each individual (230) in the subgroup of animals (130) against a set of reference codes (DR) in a source database (115) comprising a respective reference code for each individual (230) in a herd of animals in which the subgroup of animals (130) is comprised, and performing a second-stage identification of each individual (230) in the subgroup of animals (130), which second-stage identification involves matching respective registered animal-related data (Dimg) for each individual (230) in the subgroup (130) against a subset of reference data records (dr), which subset of reference data records (dr) is selected from the source database (115) based on the first-stage identification such that the subset of reference data records (dr) only contains reference data records relating to the individuals (230) in the subgroup of animals (130), characterized by the respective registered animal-related data (Dimg) for each individual (230) in the subgroup (130) comprises at least one of registered two-dimensional, 2D, image data and registered three-dimensional, 3D, image data, the subset of reference data records (dr) comprises reference image data for eachindividual (230) in the subgroup (130), wherein each reference data record is associated with an animal identity, and the method comprises: performing the second-stage identification through a process that involves matching the registered image data against the reference image data, and assigning a confirmed identity to one of the individuals in the subgroup (130) in response to a match between the registered image data and the reference image data, which confirmed identity is the animal identity associated with the matching reference data record.

13. The method according to claim 12, comprising: repeatedly performing the first-stage identification of the respective identities (IDi , ID2, ID3, ID4, IDs) of each individual in the subgroup of animals (130), and based thereon updating the subset of reference data records (dr) relating to the individuals (230) in the subgroup of animals (130).

14. The method according to claim 13, wherein the first-stage identification of the respective identities (ID1, ID2, ID3, ID4, IDs) of each individual in the subgroup of animals (130) is repeatedly performed in response to assigning the confirmed identity to one of the individuals in the subgroup (130).

15. The method according to claim 14, wherein, after having assigned the confirmed identity to said one of the individuals in the subgroup (130), the method further comprises deleting a reference data record from the subset of reference data records (dr), which deleted reference data record relates to the individual (230) to whom the confirmed identity was assigned.

16. 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 12 to 15 when the computer program (103) is run on the processing unit (101 ).

17. A non-volatile data carrier (105) containing the computer program (103) of the claim 16.

Images