Method for individually registering and identifying animals using video transmission

Abstract

The present invention pertains to the field of identifying animals by way of computer vision and by processing moving images, and more specifically relates to a method that uses streaming and video cameras to capture the muzzle of an animal, mapping the valleys and ridges in order to create a unique individual identifier, thereby facilitating registration and precise identification of the animal using muzzle biometrics.

Description

[0001] METHOD FOR REGISTERING AND INDIVIDUALLY IDENTIFYING ANIMALS THROUGH VIDEO TRANSMISSION

[0002] Technological sector of invention

[0003]

[0001] The present invention belongs to the field of animal identification by means of computer vision and motion image processing, and refers, more specifically, to a method that integrates streaming and video cameras to capture an animal's snout, mapping the valleys and ridges to create a unique individual identification, facilitating registration and accurate identification of the animal by snout biometrics.

[0004] State of the art

[0005]

[0002] Animal identification by snout biometrics is an innovative and promising technique that has gained prominence in recent years. The importance of this approach lies in its ability to provide a unique and reliable way to identify and register animals, which is crucial in several areas, such as conservation, population control and monitoring of domestic animals, guarantees in financial transactions and the origin of animals.

[0006]

[0003] Snout biometrics is based on the analysis of the unique characteristics of an animal's snout pattern, which are as distinct as fingerprints in humans. This allows for the creation of a unique biometric profile for each animal, facilitating its accurate and reliable identification.

[0007]

[0004] Among the known solutions, the use of facial recognition technology adapted for the snouts of different species stands out. Advanced algorithms are developed to analyze specific snout patterns and create a unique identification for each animal. These systems generally require the use of high-quality photographs of the animals' snouts, captured at different angles and lighting conditions to ensure accuracy and reliability.

[0008]

[0005] However, the use of photographs for snout biometrics also presents significant challenges. One of the main problems is that, in addition to the excessive time spent obtaining a single image and its processing and return, ensuring the authenticity of the images is crucial, especially in contexts where identification needs to be verified remotely. Alteration or manipulation of photographs can compromise the accuracy of the identification system through the use of photographic capture.

[0009]

[0006] Furthermore, the origin of the photographs can also be a concern, especially in cases of wild or rescued animals. Photographs obtained from unreliable sources may not accurately reflect the animal's snout pattern, which can lead to misidentification and adverse consequences.

[0010]

[0007] Another important issue is the accuracy of the facial recognition system under different environmental conditions and for different species. Variations in lighting, camera angle, and individual snout characteristics can affect the system's ability to correctly identify the animal.

[0011]

[0008] These cited problems were recognized during the application of methods described in patent documents BR102021005262-7 and BR102020014066-3 implemented by the inventors themselves, and which needed to be solved. The present invention, unlike the previous ones, does not require image-by-image capture to be processed by the method. Instead, it performs biometric mapping by video streaming and returns the response in less than 11 seconds. There is the option of using a device to record the biometrics of the animal's snout on a standard NFC RFID chip, which operates in the 13.56 MHz frequency range, inserted in an ear tag. In addition, the invention allows for the control of passage through portals for inventory purposes with a UHF RFID chip, which operates in the 865 to 928 MHz frequency range. The data is recorded in a private blockchain, guaranteeing the immutability of the information.Optionally, all identification codes (IDs) for each chip can be registered, and the manufacturing serial number of each hybrid ear tag can be tracked in a Key Management Entity, ensuring the traceability of all chips linked to the ear tag, from their production, distribution and marketing to their use by the animal.

[0009] Despite these challenges, snout biometrics remains a valuable tool for animal identification and registration. Continuous advances in individual recognition technology and image processing techniques will help improve the accuracy and reliability of these systems, making them an increasingly viable option for a variety of applications.

[0012] New Invention Details

[0013]

[0010] In this context, a method is presented that, once integrated into a system, allows for the individual identification of an animal. The invention involves a software algorithm method for animal identification by snout biometrics, using camera(s) to capture streaming and video of the snout, for the algorithm to map the location of valleys and ridges (high and low relief surfaces) to create a unique animal identification, available for animal registration and identification processes.

[0014]

[0011] The method disclosed in the invention offers a number of significant advantages. Firstly, the dynamic nature of the process allows identification to be carried out continuously and in real time, which is particularly useful in moving environments such as pastures or natural areas.

[0015]

[0012] Furthermore, continuous video capture not only enables static identification of the animal, but also serves as constantly updated proof of life. This is crucial in situations where it is necessary to verify the presence and condition of animals, such as in conservation programs or herd monitoring.

[0016]

[0013] The uniqueness of the data is guaranteed by the unique and distinct nature of each animal's snout characteristics, similar to fingerprints in humans. Customizing the identification method according to the species allows the specific characteristics of each type of animal to be considered, thus increasing the accuracy and reliability of the system.

[0017]

[0014] Furthermore, snout characteristics tend to remain relatively stable over time, ensuring the immutability of identification data. This means that even as the animal ages, the system can continue to accurately identify it over time.

[0018]

[0015] The biometric snout identification method, with video streaming capture and customized according to the species, offers a dynamic, reliable and highly accurate approach for registering and monitoring animals in a variety of contexts and applications.

[0019] Description of the attached drawings

[0020]

[0016] In order for the present invention to be fully understood and put into practice by any technician in this technological sector, it will be described in a clear, concise and sufficient manner, based on the attached drawings, which illustrate and support it, listed below:

[0021] Figure 1 represents a flowchart of the method, highlighting the registration step; Figure 2 represents a flowchart of the method, highlighting the identification step.

[0022] Detailed description of the invention.

[0023]

[0017] As shown in Figures 1 and 2, the method for animal identification by biometrics of the snout of equines, bovines, swine, caprines, ovines, felines, canines, etc., comprises the following steps: a) Positioning of the animal; b) At least one camera captures images of the animal's snout via streaming and video; c) A personal or mobile device, such as a computer or smartphone, comprises an application associated with an algorithm and database.

[0024]

[0018] The algorithm has a registration application in which:

[0025] • Streaming / video frames are decoded;

[0026] • The snout is located on the frames;

[0027] • A virtual image of the snout is formed;

[0028] • Valleys and peaks are identified, creating their coordinates; • Individual snout characteristics that differentiate one animal from another are recorded (coordinates);

[0029] • Individual characteristics are stored, assigning a unique ID to the animal;

[0030] • The information is sent to the database.

[0031]

[0019] The algorithm has an identification application in which:

[0032] • Streaming / video frames are decoded;

[0033] • The snout is located on the frames;

[0034] • A virtual image of the snout is formed;

[0035] • Valleys and peaks are identified, and their coordinates are created;

[0036] • Locate the individual characteristics (coordinates) in the database;

[0037] • A unique ID is obtained, identifying the animal.

[0038]

[0020] It is noteworthy that the camera may have a wireless or wired connection to the personal or mobile device, or it may use one belonging to a personal or mobile device.

[0039]

[0021] Furthermore, the database contains the individual biometric characteristics of each species' snout, including the animal's unique identifier.

[0040]

[0022] It is important to emphasize that the figures and descriptions provided are not intended to limit the ways in which the inventive concept now proposed can be implemented, but rather to illustrate and make understandable the conceptual innovations revealed in this solution. Thus, the descriptions and images should be interpreted in an illustrative and not limiting way, and there may be other equivalent or analogous ways of implementing the inventive concept now revealed that do not fall outside the scope of protection outlined in the proposed solution.

Claims

CLAIMS 1- METHOD FOR REGISTRATION AND INDIVIDUAL IDENTIFICATION OF ANIMALS, characterized by comprising the following steps: a) Positioning of the animal; b) At least one camera captures images of the animal's snout via streaming and video; c) A personal or mobile device, such as a computer or smartphone, comprises an application associated with an algorithm and database. 2- METHOD FOR REGISTERING AND INDIVIDUALLY IDENTIFYING ANIMALS, according to claim 1, characterized by the camera in step b) being able to have a wireless or wired connection with the personal or mobile device, or belonging to a personal or mobile device. 3- METHOD FOR REGISTERING AND INDIVIDUALLY IDENTIFYING ANIMALS, according to claim 1, characterized by the algorithm in step c) having a registration application in which: • Streaming / video frames are decoded; • The snout is located on the frames; • A virtual image of the snout is formed; • Valleys and peaks are identified, and their coordinates are created; • Individual snout characteristics that differentiate one animal from another are recorded (coordinates); • Individual characteristics are stored, assigning a unique ID to the animal; • The information is sent to the database. 4- METHOD FOR REGISTERING AND INDIVIDUALLY IDENTIFYING ANIMALS, according to claim 1, characterized by the algorithm in step c) having an identification application in which: • Streaming / video frames are decoded; • The snout is located on the frames; • A virtual image of the snout is formed; • Valleys and peaks are identified, and their coordinates are created; • Locate the individual characteristics (coordinates) in the database; • A unique ID is obtained, identifying the animal. 5- METHOD FOR REGISTERING AND INDIVIDUALLY IDENTIFYING ANIMALS, according to claim 1, characterized by the database in step c) containing the individual biometric characteristics of the snout of each species, containing the animal's unique identifier.

Images