Methods for selecting, manufacturing and / or adapting a basic muzzle shape

Abstract

A procedure for adapting a basic muzzle shape to the individual head of a dog is described with the following steps: a) Automatic selection of a suitable basic muzzle shape from a stored set of different basic shapes. The appropriate basic shape is selected from the stored set of basic shapes using a trained knowledge database that uses several digital images of the dog's head, taken from several different perspectives and depicted in the images. b) Referencing the depiction of the dog's head in the multiple digital images to a real-world scale; and c) Automatic creation of a three-dimensional data model of the dog's head from the images of the dog's head referenced to the real scale. A physical model of the dog's head is created using the three-dimensional data model of the dog's head, and the selected basic shape of the muzzle is then adapted to this physical model. Alternatively, the muzzle, adapted to the dog's head, is manufactured based on the three-dimensional data model.

Description

[0001] The invention relates to a method for selecting, manufacturing and / or adapting a basic shape of a muzzle to the individual head of a dog.

[0002] In many countries, wearing a muzzle is mandatory in certain situations and for certain dog breeds. The regulations are not uniform and vary depending on the country, state, or even municipality. Where a muzzle requirement exists, it is legally binding, and violations can be punished with fines.

[0003] On public transport such as buses and trains, a muzzle is mandatory whenever the dog is not in a transport crate. This is because public transport is crowded and can be hectic. A muzzle serves as a safety measure in case the dog reacts out of fear, insecurity, overwhelm, or simply aggression. This risk cannot be completely eliminated, even with well-socialized animals. Furthermore, wearing a muzzle may also be required in busy public areas such as pedestrian zones, event venues, or public buildings.

[0004] There is no universally applicable regulation regarding the wearing of muzzles in Germany. The individual federal states, and sometimes even the municipalities, decide on the relevant guidelines.

[0005] However, the muzzle requirement applies particularly to dogs on the list of dangerous breeds in the respective federal states.

[0006] The categorization of breeds as dangerous dogs is generally based on statistical data regarding dog bite incidents. Here is a selection of dog breeds that are subject to muzzle requirements in certain German states: • American Staffordshire Terriers • American Pit Bull Terrier • Bull Terrier • Staffordshire Bull Terrier • Rottweiler • Tosa Inu.

[0007] In most cases, mixed-breed dogs are also affected by the relevant regulations.

[0008] Dogs can also be classified as dangerous and therefore subject to muzzle requirements if they have already become conspicuous or exhibit increased aggressive behavior.

[0009] Owners of "dangerous dogs" and "dogs of certain breeds" for which a general muzzle requirement applies can be exempted by applying for a special permit and providing proof of a medical indication or a positive temperament test. A medical indication means that the dog suffers particular distress due to an illness when wearing a muzzle. This must be examined and certified by a veterinarian. A prerequisite for an official exemption is that the dog poses no danger. If this cannot be clearly ruled out, an additional official veterinary assessment may be required. The temperament test is usually conducted with the dog and the owner in person. In most cases, the animal must be at least 15 months old. An official veterinarian or a recognized expert examines the animal against specific behavioral criteria.If the test is passed, a positive certificate is issued, which can be submitted to the administrative authority to apply for an exemption. However, an exemption does not exempt the dog from the muzzle requirement everywhere. In locations where muzzles are mandatory for all dogs, even animals with an exemption must wear one.

[0010] At the same time, the number of dogs imported into Germany from abroad continues to rise, and we have more and more dogs that are misunderstood, not adequately supported and trained, and therefore become "problematic".

[0011] Animal shelters are full, and muzzles are a necessary tool to be able to interact and work with these dogs at all.

[0012] Besides owners of dogs with unusual behavior, every dog ​​owner should have a muzzle at home and their dog accustomed to wearing one. Especially at the vet, and particularly if the dog is in severe pain, it can behave unpredictably and, in some cases, bite. For everyone's safety, a muzzle is essential in these situations, and ideally, the dog should already be used to wearing one.

[0013] Another reason for using a muzzle is the danger of poisoned bait, which can be virtually eliminated with a properly fitted wire muzzle and a bait guard. The "poison bait nets" offered are unsuitable because they do not allow the dog sufficient room to pant, making their use unnatural.

[0014] Dog breeds – differentiated by sex and age, and thus also by varying body and head sizes within the breed – require an individual muzzle shape adapted to their specific head form. This ensures the muzzle fulfills its purpose and meets the requirements of humane wear. Currently, numerous basic metal muzzle shapes are offered by several national and international companies, allowing for an initial assessment of breed and size. These standard shapes, however, usually require adjustment to the individual head shape and size, which is currently done on the live animal. This involves placing the supposedly suitable muzzle on the animal and manually achieving an optimal and dog-friendly fit by bending the muzzle and readjusting / replacing the nose pad, neck strap, and additionally attaching the throat strap.

[0015] DE 19 76 851 U discloses a muzzle for dogs, which is formed from individual straps that are at least partially connected to one another. For adjustment purposes, the straps are each made of a flexible, endlessly manufactured and cut-to-length material band of plastic-coated fabric.

[0016] DD 15 698 A5 discloses a muzzle for dogs which incorporates carrying straps in interconnected side pieces, a headband connected to the side pieces with guides, and loops attached at the rear. The muzzle is adjustable to different sizes by means of the carrying straps.

[0017] US 2015 / 0 257 368 A1 describes a heat-malleable semi-finished product for adapting a muzzle.

[0018] In KR 10 2022 0 022 209 A a clip-on adjustable muzzle is disclosed.

[0019] 3D software. In: 3D-Medico. Last edited: January 16, 2025, 11:20 UTC. URL: https: / / web.archive.org / web / 20250116112049 / https: / / 2025.3d-medico.de / 3d-software.html#content5-1u [accessed October 22, 2025] discloses AI-supported online 3D orthotics software for the production of individually customized orthopedic aids using 3D printing.

[0020] 3D Busts. In: youlittle. Last edited on 8 November 2024, 22:12 UTC. URL: https: / / web.archive.org / web / 20241108221202 / https: / / youlittle.com / 3d-figuren / 3d-busten / [accessed on 22.10.2025] reveals the 3D printing of human 3D busts based on a 3D face scan.

[0021] Limuzz Basic. In: Limuzz. Last edited on December 7, 2024, at 00:08 UTC. URL: https: / / web.archive.org / web / 20241207000806 / https: / / limuzz.com / produkt / limuzzbasic / [accessed on October 22, 2025] reveals a custom-made muzzle for dogs, individually manufactured based on the specified height, length, circumference, and width of the closed muzzle.

[0022] Face masks in sports. In: Manfred Klein Medical Supply Store. Last edited: September 9, 2024, 2:40 PM UTC. URL: https: / / web.archive.org / web / 20240909144011 / https: / / www.klein-sanitaetshaus.de / orthop%C3%A4die-technik / gesichtsschutzmaske / [accessed October 22, 2025] reveals face masks that are individually adapted using 3D printing as face protection for people in sports.

[0023] Manually fitting a muzzle requires considerable experience to select a suitable metal muzzle with an appropriate basic shape. A wide variety of basic shapes are available from numerous suppliers. The fitting process itself always carries risks, as the unfamiliar situation of constantly putting the muzzle on and taking it off, combined with the additional adjustments made by the consultant in the dog's biting area, can lead to injuries due to unforeseen reactions from the dog.

[0024] Usually, an adjustment is necessary because the fit of a standard muzzle is not optimal for the dog. The muzzle is too narrow, too wide, or the distance between the muzzle and the nose or eyes is too short or too far. Often, the dog can pull the ill-fitting metal muzzle off its head by scratching, making safe use in public impossible. On-site adjustments, combined with repeated putting on and taking off, can lead to the dog associating the first time it wears a muzzle with a negative experience, making it difficult for the dog to get used to the situation and to accept the muzzle. A significant investment of time and expertise is required to determine the correct measurements of the dog's head, find a local specialist, and schedule an appointment.The necessary travel time for arrival and departure must be taken into account: Fitting typically requires approximately 30 minutes. Furthermore, it is still not guaranteed that the appropriate basic muzzle will be available on site, and further scheduling may be necessary.

[0025] The object of the present invention is therefore to create an improved method for adapting a basic shape of a muzzle to the individual head of a dog.

[0026] The problem is solved by the method with the features of claim 1. Advantageous embodiments are described in the dependent claims.

[0027] The procedure for adapting a basic muzzle shape to the individual head of a dog involves the following steps: a) Automatic selection of one or more suitable basic muzzle shapes from a stored set of different basic shapes. The appropriate basic shape is selected from the stored set of basic shapes using a trained knowledge database that compiles multiple digital images of the dog's head, photographed from several different perspectives and depicted in the images. b) Referencing the image of the dog's head in the multiple digital images to a real scale; c) automatic creation of a three-dimensional data model of the dog's head from the images of the dog's head referenced to the real scale; and d1) Creating a real model of the dog's head using the three-dimensional data model of the dog's head as well as d2) Adapting the selected basic shape of the muzzle to the manufactured real model.

[0028] Currently, only recommendations for an optimal metal muzzle exist, based on available measurements, experience, and the suitability of different dog breeds and genders. Various manufacturers provide recommendations on their websites, including additional information regarding size, etc., which can be used as guidelines. However, a standardized definition and corresponding size specifications or recommendations across all suppliers do not yet exist. Furthermore, a custom fit is not available and must always be tailored to the individual dog.

[0029] The inventive method allows for an optimal, manufacturer-neutral selection for each individual dog, regardless of breed or the various suppliers of base models, thus minimizing the need for subsequent adjustments. The automated production of a three-dimensional model of the dog's head allows for adjustments to be made to this model rather than to the live dog. This makes the fitting process simpler, more humane, and safer. Furthermore, the model can later be displayed by the user as a decorative item or keepsake.

[0030] By transferring and using the process via an app on a smartphone, the application can be scaled through appropriate licensing, which can be carried out independently by the owner or by other licensees (companies, dog chains, platforms, organizations, dog schools, etc.).

[0031] The order of the steps can be varied if necessary. For example, step b) can be performed before step a).

[0032] The physical model can be manufactured automatically using additive or subtractive manufacturing. Suitable methods include 3D printing, such as selective laser sintering / melting (SLS / SLM), fused deposition modeling (FDM), stereolithography (STL / SLA), and similar techniques.

[0033] It is possible to automatically identify the adaptations of the selected basic shape of the muzzle to the modeled dog's head using a trained neural network, which calculates the adaptation of the three-dimensional shape of the selected basic model to the shape of the dog's head depicted in the digital images from the multiple digital images of the dog's head, which is taken and depicted in several different perspectives, and the stored three-dimensional shape of the selected basic model, and outputs it for identification.

[0034] For this purpose, the entered digital images of the dog's head can be referenced to a real-world scale. It is advantageous to include a scale in the images themselves. Alternatively, characteristic pixels and image features can be referenced to each other to automatically determine the scale. This can be achieved using grid patterns in the background or by taking multiple images from different perspectives with a known optical alignment of the multiple lenses used.

[0035] The identified adjustments can be displayed in at least one view for manual modification of the basic shape. This provides the user with the necessary instructions for making the adjustments. The modified muzzle can then be photographed, and the multiple images of the modified muzzle can be projected onto the created three-dimensional model of the dog's head using image processing logic. This allows verification that the modifications are optimally suited to the dog's head.

[0036] The identified adjustments can be output to a file and transferred to a manufacturing unit for automatic adjustment of the basic shape or for automatic production of the adjusted basic shape with the manufacturing unit.

[0037] It is possible to automatically extract characteristic features of the dog from the automatically generated three-dimensional data model of the dog's head using a trained knowledge model and to transfer these characteristics, along with data from the dog owner, to automatically request an insurance quote for the dog. The characteristic features of the dog, automatically extracted through the recordings using the AI ​​application and the reference database, can be combined with user data using a user-provided software application that handles the execution and transfer of this data.

[0038] The evaluation of the recordings is supported and transferred to an external service provider, who automatically creates an insurance contract offer tailored to the respective dog and user and automatically transmits it to the user.

[0039] The procedure can be carried out automatically using a computer program, wherein the computer program includes instructions that cause a computer, upon execution of the instructions, to perform the procedure steps a) to c) and to forward the three-dimensional data model of the dog's head to produce a real model of the dog's head in step d1) for the adaptation of a basic model of a muzzle or in step e) for the production of an adapted muzzle.

[0040] The invention is explained in more detail below with reference to an exemplary embodiment and the accompanying drawings. These show: Fig. 1 - Sketch of the procedure for adapting a basic muzzle shape to the individual head of a dog; Fig. 2 - Flowchart of steps a) to e) of the procedure.

[0041] Fig. Figure 1 shows a sketch of the procedure for adapting a basic muzzle shape to the individual head of a dog.

[0042] In a first step 1), the dog owner or user N typically creates one or more digital photos F and / or videos of the dog's head and muzzle from several perspectives, especially side view, top view, and front view. Optionally, a photo or video of the entire dog can also be taken.

[0043] To reference the photos or videos to a real scale (dimensions, length, width, height), a parallel shot of the head can be taken with a scale M (measuring stick, grid pattern in the background, algorithms using a multi-lens shot via a suitable recording device such as an iPhone). The photos (F) or videos, for example, taken using a software application (APP), are transferred by the user or their app in step 2) to a central service server where an AI application (Al) is running. The AI ​​application (Al) is configured to automatically merge the photos (F) and / or videos using a trained image-based network or knowledge model and create a three-dimensional model (3D-M) of the dog's head depicted in the photos.

[0044] This can be achieved by rectifying the captured images of the dog's head and calibrating the rectified images. Such rectification allows the captured images to be adjusted to a given perspective.

[0045] For the AI ​​application Al, a generative neural network Al is trained based on existing datasets of images of dog heads, each assigned to a predefined 3D data model 3D-M of the dog head. Using images of the dog head taken by the user and the trained neural network, a file D is created that describes the 3D head model. The three-dimensional data model 3D-M can be edited or created manually, semi-automatically, or automatically by the user using a computer program running on a computer. The file has a 3D model file format compatible with the computer program and enables virtual representation.

[0046] In step 3), a dog head model H is created from file D using a milling machine from a blank material - plaster, wood, metal or manufacturing via 3D printing.

[0047] The adaptation of a muzzle takes place in step 4) using data from a comprehensive reference database R of existing basic shapes of metal muzzles, which serves as the basis for an automated, knowledge-based AI application Al. The reference database R can be continuously updated, for example, with new releases. The reference database R can also contain reference photos or videos of known dog heads for the various dog breeds.

[0048] A computer-aided automatic analysis and selection of an optimal basic shape (best fit, least deviation of dog's snout, basic shape, quality features of optimal, dog-friendly wearing comfort) for a muzzle is carried out from the reference database R, taking into account the specific dog's snout from the respective recordings F or the three-dimensional wire model 3D-M derived from them using the AI ​​application.

[0049] Depending on the reason for wearing it (short-term application with a compact fit or long-term wear with appropriate depth), a second option for a variant of the muzzle can be specified.

[0050] In step 5), the necessary manual adjustments to the selected basic muzzle shape are automatically identified using computer assistance to fit the specific dog's head or snout. For example, with a metal muzzle, the necessary bends at specific points, surfaces, and differences are defined and identified in terms of location, extent, and dimension in a file referenced to the calculated 3D data model. The adjustment of the basic muzzle model in step 6) can be carried out using the following methods, for example: - manual adjustment by a human using the calculated and displayed corrections; - Insertion and fixing of the muzzle base model into a mechanical device which automatically performs the identified rework based on the calculated 3D file D; - Use of the created model (plaster, wood, 3D print, etc.) as a reference for the dog's snout and manual adjustment of the muzzle's basic shape to the model by a person; - virtual representation of the dog's head / snout and a selected basic model as a basis for manual adjustment by a human; - Analysis and target-actual comparison of the current manual adjustments during the process and dynamic display of the achieved result via an imaging procedure through repeated or continuous photo or video recording of the modified muzzle - also in several processing phases.

[0051] The photos of the dog's head, snout and the dog itself can be taken using an app and a smartphone, with the location and time of the recording also being recorded alongside the photos.

[0052] The evaluation and calculation of the 3D files D and the resulting output can be performed using either a cloud-based algorithm or on the smartphone itself using appropriate algorithms.

[0053] In addition to the processes described, the application can also enable coupling with other applications P in step 7). These include: • Transmission and storage of the contact details of the dog owner or keeper; • Configuration of a customer / client process for ordering, paying for and billing a corresponding service; • Further services in connection with muzzle fitting, such as insurance / liability for dogs / puppies, which are offered a special rate based on the use of an optimally fitted metal muzzle, derived from the three-dimensional data model 3D-M or characteristic features of the selected basic muzzle shape and its associated properties. In optional step 8), contract offers can be automatically generated by computer (possibly using a knowledge-based database or a trained neural network) based on the transmitted contact details and characteristic features of the dog determined using the AI ​​application Al, as well as any data entered about the dog, and then transmitted to the user N.

[0054] Fig.Figure 2 shows a flowchart of the procedure for adapting a basic muzzle shape to the individual head of a dog, including the following steps: a) Automatic selection of a suitable basic muzzle shape from a stored set of different basic shapes using a trained knowledge database that selects the appropriate basic shape from the stored set of basic shapes based on several digital images of the dog's head, which is taken and depicted from several different perspectives; b) Referencing the image of the dog's head in the multiple digital images to a real scale; c) automatic creation of a three-dimensional data model of the dog's head from the images of the dog's head referenced to the real scale; and d1) Creating a real model of the dog's head using the three-dimensional data model of the dog's head as well as d2) Adapting the selected basic shape of the muzzle to the manufactured real model.

[0055] The sequence of steps can vary. For example, referencing the image of the dog's head in step b) can occur before automatically selecting the appropriate basic shape of the muzzle.

Claims

[1] Method for adapting a basic muzzle shape to the individual head of a dog, including the following steps: a) Automatic selection of a suitable basic muzzle shape from a stored set of different basic shapes. The appropriate basic shape is selected from the stored set of basic shapes using a trained knowledge database that uses several digital images of the dog's head, photographed from several different perspectives and depicted in the images. b) Referencing the image of the dog's head in the multiple digital images to a real scale; c) automatic creation of a three-dimensional data model of the dog's head from the images of the dog's head referenced to the real scale; and d1) Creating a real model of the dog's head using the three-dimensional data model of the dog's head as well as d2) Adapting the selected basic shape of the muzzle to the manufactured real model. [2] Method according to claim 1, characterized by , that step b) is performed before step a). [3] Method according to claim 1 or 2, characterized by that the production of the real model is done automatically using additive manufacturing or subtractive manufacturing. [4] Method according to any one of claims 1 to 3, characterized byAutomatic identification of the adaptations of the selected basic shape of the muzzle to the modeled dog's head using a trained neural network, which calculates and outputs the adaptation of the three-dimensional shape of the selected basic shape to the shape of the dog's head depicted in the digital images from the multiple digital images of the dog's head, which is taken from several different perspectives and depicted in the images, and the stored three-dimensional shape of the selected basic shape for identification. [5] Method according to claim 4, characterized by , that the identified adjustments are output in at least one representation for manual adjustment of the basic form. [6] Method according to claim 4, characterized by, that the identified adjustments are output in a file and transferred to a manufacturing unit for automatic adjustment of the selected basic shape of the muzzle or for automatic production of the adjusted muzzle with the manufacturing unit. [7] Computer program comprising commands which, when executed by a computer, cause it to perform the process steps a) to c) according to claim 1 and to forward the three-dimensional data model of the dog's head for the production of a real model of the dog's head in step d1) or in step e) for the production of a fitted muzzle.

Images