Device for identifying and activating high speed ovens
The integration of a USB camera and SoM with AI in speed ovens automates food recognition, addressing inefficiencies by providing precise and rapid cooking parameter selection, thus enhancing operational efficiency and reducing energy consumption.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- PRATICA PROD
- Filing Date
- 2025-12-15
- Publication Date
- 2026-07-02
AI Technical Summary
Existing speed ovens lack integrated food recognition systems that automate the cooking process, leading to inefficient manual operation and potential errors in recipe selection, which increases preparation time and energy consumption.
A device integrating a USB camera and a System on Module (SoM) with edge artificial intelligence to capture food images, process them using convolutional neural networks, and automatically select the appropriate cooking parameters based on a trained database, eliminating the need for user intervention.
The system achieves high-precision food recognition with reduced response time, minimizing errors and optimizing cooking processes by automating recipe selection and reducing preparation time.
Smart Images

Figure BR2025050583_02072026_PF_FP_ABST
Abstract
Description
"DEVICE FOR IDENTIFICATION AND ACTIVATION OF SPEED OVENS"
[0001] This invention relates to capturing a high-resolution color image of food as it enters the equipment, designed to capture its image and process data on the type of food. With this information, the oven is automatically programmed to cook the food, without the need for any user intervention, and belongs to the field of ovens used for cooking food.
[0002] This device was developed to save time and make the oven user's job easier, eliminating the need to press buttons or even know the correct program for that type of food.
[0003] With the growing demand for operational efficiency and reduced energy consumption, the market for food cooking ovens, especially for the quick preparation of snacks and savory pastries, has been incorporating technological solutions that seek to replace traditional convection ovens. This evolution meets the demands for greater speed, uniformity in preparation, and sustainability in energy consumption.
[0004] The application of more advanced technologies for optimizing thermal processes has allowed the food industry to develop equipment that not only ensures greater energy efficiency, but also guarantees better operational control.
[0005] The installation of such devices, including a camera, is designed to address the need for automation and precision in food recognition in Speed Ovens, offering faster and more even cooking without the user having to program the temperature and cooking time, significantly reducing the time required to prepare these foods.
[0006] The items and actors mentioned above are essential in high-demand environments, such as commercial / industrial kitchens, fast food restaurants, and bakeries. The aforementioned factors unequivocally justify granting the exclusive right constitutionally guaranteed to invention patents, in accordance with the precepts of the legal norm. STATE OF THE ART
[0007] The market for fast cooking ovens, especially Speed Ovens, has established itself as an essential solution in environments where cooking time is a critical factor, such as industrial kitchens, fast food services, and commercial establishments that demand high operational efficiency. These appliances combine advanced cooking technologies, such as forced convection, microwaves, and infrared, to guarantee fast and consistent results.
[0004] With increasing environmental concern and the need to reduce energy consumption, the furnace industry has been incorporating innovations that not only meet sustainability requirements but also improve operational performance. These innovations aim to align with environmentally sound policies and promote greater energy efficiency, contributing to cost reduction and the adoption of more economical and ecological practices.
[0008] Preliminary research was conducted to identify similar food recognition systems. While generic computer vision solutions exist, no complete integration with speed ovens using dedicated hardware such as SoM (System on Module) was found. Specifically within the speed oven category, there is no solution related to this type of application or technology. Let's look at: a. AU2016321324B2 Australia - In-oven camera, abstract: Several embodiments include a cooking appliance. The cooking appliance may include one or more heating elements; a cooking chamber; and a camera attached to the interior of the chamber. In some embodiments, the cooking chamber prevents any visible light from escaping the chamber (e.g., the cooking chamber has no windows). In some embodiments, the heating elements are controlled by a computing device in the cooking appliance.In some embodiments, the camera output is used to adjust the heating pattern of the heating elements: Describes a furnace with an internal camera for adjusting the heating. b. US20240094060A1 United States - In-oven camera and computer vision systems and methods - Abstract: The systems and methods include a cooking apparatus comprising a heating element disposed within a cooking chamber and operable to selectively emit waves at any one of a plurality of powers and / or peak wavelengths, an operable camera to capture an image of the cooking chamber, and an operable computing device to supply power to the heating element to vary the power and / or peak wavelength of the emitted waves and generate heat within the cooking chamber and instruct the camera to capture the image when the heating element is emitting at a stabilized power and / or peak wavelength. The computing device is operable to generate an adjusted captured image by adjusting the captured image with respect to the stabilized power and / or peak wavelength.The computing device comprises operable feedback components to receive the captured adjusted image, extract features, and analyze one or more features to determine an event, property, measurement, and / or status. c. EUA11828658B2 United States - In-oven camera and computer vision systems and methods: Describes systems and methods that include a cooking apparatus comprising a heating element disposed within a cooking chamber, operable to selectively emit waves with different power levels and / or peak wavelengths. A camera is configured to capture an image of the cooking chamber, and a computing device is operable to supply power to the heating element, adjusting the power and / or peak wavelength of the emitted waves, generating heat within the chamber, and instructing the camera to capture an image when the heating element is operating at a stabilized power and / or peak wavelength. The computing device can generate an adjusted captured image by calibrating the captured image relative to the stabilized power and / or peak wavelength.
[0009] The model proposed in the patent document described herein differs from the models above by offering high precision, reduced response time, and direct integration with the oven control. It is also optimized for real-world kitchen use conditions.
[0010] In all documents found in the prior art, the solution proposed in this document is not presented. However, despite technological advances in the Speed Oven segment, the market continues to demand solutions that increase energy efficiency, optimize cooking processes, and guarantee reduced preparation time without compromising food quality. The present invention fits into this context, seeking to overcome current limitations and offer significant improvements in the performance and energy savings of these appliances. Summary of the invention
[0011] It was with the aim of solving this problem that, after numerous research and studies, the inventor created and developed the object of this patent, designing a device to capture an image of the food and process the captured data through software that, with this information, automatically programs the oven to cook the food.
[0012] The proposed system is designed to address the need for automation and precision in food recognition in speed ovens. The patent seeks to protect the technological innovation that integrates dedicated hardware and artificial intelligence algorithms.
[0013] The proposed model solves problems of inefficient manual operation, offering automated solutions for food identification and recipe application. This technology promotes efficiency by reducing errors and optimizing processes.
[0014] Unlike other oven categories, the use of speed ovens is characterized by their speed. This type of equipment allows for quick preparations, often in less than 2 minutes, in fast food restaurants, cafeterias, convenience stores, bakeries, and other establishments where demand is often very high and customer waiting time needs to be minimal. In this sense, solutions that streamline equipment operation and consequently reduce customer waiting time are very important. BRIEF DESCRIPTION OF THE INVENTION DRAWINGS
[0015] To complement the present description in order to obtain a better understanding of the characteristics of the present invention and in accordance with a preferred practical embodiment thereof, a set of drawings is attached to the description, where, in an exemplary, though not limiting, manner, the following is represented: a. Figure 1 - Shows a front view of the oven with the door open and food inserted. b. Figure 2 - Shows a side view of the oven and a figure representing the insertion of food into the oven with the image captured by the camera. c. Figure 3 - Shows a graph of information processing by AI and the corresponding recipe for food processing. TECHNICAL DESCRIPTION OF THE INVENTION
[0016] The present invention solves the problem of erroneous or time-consuming manual identification of food placed in the oven, eliminating the user's dependence on selecting the appropriate recipe according to the respective pictures. Furthermore, it minimizes oven setup errors and reduces preparation time.
[0017] According to Fig. 1, the present model uses a USB camera (1) to capture images of the food (2) and a dedicated electronic module, based on a high-processing SoM (System on Module) with advanced features for implementing edge artificial intelligence, to perform the processing of the food images (2). The solution employs convolutional neural networks to compare the captured images with a mathematical model (3) formed from training performed with a food image bank, according to Fig. 3. Upon reaching a similarity index (4) greater than 80%, the system automatically loads the corresponding recipe (5).
[0018] The present device is mainly composed of the following elements: USB camera (1) for capturing images of food (2), processing module with NPU for AI acceleration, as shown in figure 3, embedded inference software in Linux (3), similarity check of images with the food database (4), user panel (8) for manual configuration and monitoring.
[0019] The system uses artificial intelligence (AI) to prepare food automatically and, integrated with a recipe database, to perform the recognition and optimized preparation of different types of food without the need for human intervention. The process begins with the system capturing visual information and / or characteristics of the food (2) through a camera (1) inserted in the equipment. Through the information processing module, the system employs AI algorithms to identify the specific type of food (3) based on patterns previously stored in its database.
[0020] Once the food is recognized (4), the system automatically accesses the ideal recipe for preparing that item (5), including parameters such as temperature, cooking time, specific preparation methods, and any additional steps required. Based on this data, the system automatically adjusts the components of the cooking appliance (6), such as heating elements, ventilation, and timers, ensuring that the food is prepared accurately and consistently.
[0021] The device software is based on trained neural networks. The training and model generation system optimized for embedded systems and embedded Linux environment recognizes images of food (2) and automatically provides a recipe (5) for the product to be prepared, without user intervention. The respective software (3), applied to the invention discussed here, can be applied in commercial, industrial and residential kitchens that use speed ovens, promoting automation and operational efficiency.
[0022] The system starts capturing images when the oven door is opened (7). While the user is loading the food (2), the camera (1) records a video of the food (2) and sends the data to the processing module (3). Then, in about 2 seconds, the system identifies and automatically recognizes (2) the food with an accuracy above 80%. Then, it downloads the corresponding recipe (5) or, if it does not exist in the database, the system displays suggestions on the user panel (8), where they can add a new recipe.
[0023] We would like to highlight that the device features integrated hardware and embedded software and is compatible with different types of food and lighting conditions.
[0024] The present invention exhibits broad compatibility with commercial and industrial operations, meeting the demands of markets that require high efficiency, optimized performance, and operational reliability. Furthermore, its characteristics allow for application in residential, commercial, or industrial environments, making it a versatile and adaptable solution for different usage scenarios. This flexibility expands the scope of the invention, meeting the stringent requirements of industrial environments as well as the practical and functional needs of home users.
[0025] The image recognition system for speed ovens is an innovative solution that automates the food identification process and optimizes oven usage. The user simply places the food in the oven, which then identifies it and activates the pre-defined cooking method for that dish. It offers clear benefits in terms of efficiency, accuracy, and user experience.
[0026] If there is no recipe that matches the one entered in the oven, the technology triggers a warning so that the user can add the recipe to the user panel (8). Then, with this information included, the AI is trained to make the addition taking into account the user's operations.
[0027] It is understood that when the present invention is put into practice, modifications may be introduced with regard to certain details of construction and form, without departing from the fundamental principles that are clearly substantiated in the claims, it being understood that the terminology used was for descriptive purposes and not for limitation.
Claims
CLAIMS 1. "DEVICE FOR IDENTIFICATION AND ACTIVATION OF SPEED OVENS CATEGORY", refers to equipment for preparing fast food such as so-called speed ovens, characterized by having a device, such as a camera (1) for recognizing images of the food (2) through a computer system based on artificial intelligence (3) that recognizes the food and downloads a recipe (5) that includes preparation parameters (6) such as heating elements, temperature, ventilation inside the device, humidity, cooking time through timers and other specific preparation methods without user interference.
2. "DEVICE FOR IDENTIFICATION AND ACTIVATION OF SPEED OVENS CATEGORY", refers to equipment for preparing fast food such as so-called speed ovens, characterized in that if the defined recipe is not found in its database, the system informs the user through the panel (8) of this condition and the user can enter new information for cooking the product which can be included in the database.