Gesture recognition system

Abstract

A gesture recognition system comprises at least one sensor (1) and at least one controller (2) comprising at least one processor (3). The at least one controller (2) further comprises software. A power supply is configured to power the at least one sensor (1) and the at least one controller (2). The gesture recognition system further comprises at least one device and / or at least one apparatus and an effector configured to activate the at least one device and / or the at least one apparatus. The at least one sensor (1) is located within a usage area of a static or mobile environment or entity and is configured to communicate with the at least one controller (2). The number of the at least one sensor (1) is at least one less than the number of the at least one device and / or at least one apparatus.

Description

Technical Field

[0001] The present invention relates to a posture recognition system that can be used in any area of ​​use, such as buildings and / or vehicles, such as commercial and / or military aircraft, in which the device can be activated by, for example, passengers and / or crew members.

[0002] For example, such a posture recognition system can be used for cockpit device activation and / or equipment activation in vehicles (e.g., aircraft). Background Technology

[0003] For example, an aircraft cockpit may contain multiple individual devices and / or individual equipment that can be triggered and activated by passengers and / or crew members. Such devices and / or equipment typically consist of at least one sensor, which is either a mechanical or electrical sensor, that triggers actuators to activate the function of each individual device and / or each individual piece of equipment.

[0004] For example, such devices and / or equipment may include reading lights integrated into passenger service units, or infrared sensors designed for faucets (e.g., faucets built into restrooms).

[0005] Issues related to existing technology Most individual devices and / or individual equipment existing in the prior art are independent systems that do not communicate with each other and / or with the cockpit management system, and are not even capable of such communication. Current products in the prior art include at least one sensor for each individual device and / or each individual equipment for each activated function. These sensors typically provide only simple and limited information, which is simple and limited binary information (e.g., mechanical pressure threshold such as a mechanical switch).

[0006] Furthermore, the existing individual devices and / or these individual devices lack the ability to store any data about their use (e.g., activation frequency), which could be further analyzed for design improvements and / or predictive maintenance.

[0007] Technical objective of this invention The main technical problem to be solved by this invention is: - Reduce the number of individual sensors in existing technologies. - Collect and centralize data within a single system. - To achieve more accurate activation of devices and / or equipment based on the true intent of the target, thereby further improving the experience and comfort of passengers and crew. - Collect the output data of artificial intelligence algorithms to analyze and further improve the product design of this invention in terms of economic and ecological aspects, such as reducing unintentional activation of faucets and / or reading lights. - Compared to existing limited sensor information (e.g., the presence or absence of objects), more accurate data can be obtained through further information about the behavior and movement of passengers and / or crew members, and - Improve hygiene and / or maintenance by using non-contact activation and minimizing the use of conventional mechanical switches.

[0008] Summary of the Invention The posture recognition system according to the present invention can achieve the above-mentioned objectives by using the technical features disclosed in these application documents.

[0009] Pose recognition systems can be used in any type of static or moving environment or entity, as further described below.

[0010] Throughout this application, such an environment or entity will be described as the area of ​​use.

[0011] The gesture recognition system can be used in any area of ​​use, such as vehicles, such as aircraft, trains, buses, trucks, ships, or automobiles; more particularly, for passenger vehicles that transport passengers, such as passenger planes, passenger trains, passenger buses, passenger trucks, passenger ships, or buses; and even more particularly, for corresponding toilets, cabins, passenger and / or crew rest areas, galleries and / or kitchenettes on said passenger vehicles, such as toilets, cabins and / or galleries of aircraft, toilets, cabins, galleries and / or kitchenettes of trains, toilets, cabins, galleries and / or kitchenettes of buses, toilets, cabins, galleries and / or kitchenettes of trucks, toilets, cabins, galleries and / or kitchenettes of ships, or toilets, cabins, galleries and / or kitchenettes of recreational vehicles (RVs).

[0012] The gesture recognition system can also be used in any area of ​​use such as buildings, houses, apartments, or fixtures, such as toilets, cabins, kitchens, and / or kitchenettes in buildings, houses, apartments, or fixtures.

[0013] Examples of equipment or devices include contactless, push-button, mechanical, electromechanical faucets and / or lights, soap dispensers or water dispensers in toilet or kitchen cubicles, or disinfectant dispensers located in aircraft cabins.

[0014] Examples of equipment or devices may also include cabin crew call units, seat tilting mechanisms, gasper switches, and / or overhead storage compartment opening / locking mechanisms.

[0015] The terms “equipment” and “device” are used interchangeably; however, their meanings are understood according to the conventional meaning of each word and should not be considered synonyms.

[0016] Technical solution and beneficial effects of the present invention This invention aims to replace and / or reduce the number of existing sensors consisting of multiple individual devices and / or individual equipment by providing at least one posture recognition system capable of analyzing and identifying the postures of passengers and / or crew members, and then triggering at least one actuator to activate the corresponding individual device and / or corresponding individual equipment. The posture recognition system of this invention is intended to be able to be integrated into locations such as those described above, for example, any cockpit area such as commercial and / or military aircraft.

[0017] The system of the present invention can further centralize data management to simultaneously sense and trigger multiple devices and / or equipment, thereby reducing the number of individual sensors on the target device and / or equipment. Examples may include the operation of reading lights in the cockpit, seat tilting function, and / or activation of a faucet in the toilet, but the invention is not limited thereto, but relates to any device and / or equipment in the aforementioned locations.

[0018] Furthermore, this invention for gesture recognition can distinguish between intentional and unintentional activation caused by gestures. For example, this invention can distinguish whether activation via a passenger's head is intentional or unintentional, whether it is activating his or her reading light or personal items (such as a handbag), or activating a faucet in a toilet.

[0019] This can be achieved by training the AI ​​with possible unintentional actions that should not activate the device and / or equipment.

[0020] The posture recognition system of the present invention can integrate an artificial intelligence model that can accurately analyze and predict the intentions of passengers and / or crew members by analyzing their motion and position data (collected by one or more of at least one sensor). At least one controller can trigger signals based on the output of the artificial intelligence algorithm to simultaneously activate one and / or more cabin devices and / or one and / or more equipment.

[0021] By implementing this invention, the following aspects can be improved: - Reduce the number of individual sensors in existing technologies. - Collect and centralize data within a single system. - Activating devices more precisely based on the object's true intent can further improve passenger experience and comfort. - Collect the output data of artificial intelligence algorithms to analyze and further improve the product design of this invention in terms of economic and ecological aspects, such as reducing unintentional activation of faucets and / or reading lights. - Compared to existing limited sensor information (e.g., detecting the presence or absence of objects), more accurate data can be obtained through further information about the behavior and movement of passengers and / or crew members, and - Improve hygiene and / or maintenance by using non-contact activation and reducing existing mechanical switches.

[0022] Efforts have been made to train artificial intelligence models to accurately classify input data from at least one sensor based on the object’s intent and to eliminate unintentional motions for the activation of devices and / or equipment, and to implement the model in a system to activate various devices and / or equipment (e.g., toilet devices).

[0023] Compared to existing solutions in the prior art, the artificial intelligence model of the present invention can identify more complex situations and thus more accurately predict the intentions of passengers and / or crew members, and distinguish between intentional and unintentional activation situations.

[0024] Furthermore, this invention enables the collection and recording of sensor data, as well as the data analysis process. Attached Figure Description

[0025] Figure 1 An architecture for a gesture recognition system for at least one of the sensors is shown.

[0026] Figure 2 An example of a gesture recognition system architecture in a toilet environment according to an embodiment of the present invention is shown.

[0027] Figure 3 An example of data provided by sensors in a gesture recognition system architecture is shown.

[0028] Figure 4 An example of the architecture of a gesture recognition system in an overhead cockpit is shown.

[0029] Detailed Description of the Embodiments (see attached figures) Figure 1 The architecture of the posture recognition system for a sensor (1) of the present invention is shown.

[0030] The movement and / or position of passengers and / or crew members are identified by sensor (1). Posture data is transmitted from sensor (1) to posture controller (2), which includes corresponding software and / or artificial intelligence functions. The posture data is analyzed and posture recognition is performed, which can distinguish between intentional and unintentional activation caused by posture.

[0031] Based on the posture recognition, the posture controller (2) can trigger signals for the corresponding device and / or the corresponding equipment.

[0032] Figure 2 An example of the posture recognition system of the present invention in a toilet environment is shown. A sensor (1) (e.g., located at position 1) and a controller (2) (e.g., located at position 2 (e.g., above the ceiling (4), e.g., located at position 4)) analyze the movement of a passenger or crew member (7) (e.g., located at position 7). The analysis is directed to a device and / or equipment, such as a trash baffle (8), which may be located at position 8.

[0033] The gesture recognition system includes at least one sensor (1) (e.g., located at position 1) and at least one controller (2) (e.g., located at position 2). The at least one controller (2) includes a processor (3) (e.g., located at position 3). The processor (3) operates using software employing artificial intelligence algorithms.

[0034] The type of at least one sensor (1) is not limited to a specific type, such as an optical sensor (e.g., such as an optical camera, lidar, time-of-flight sensor or camera), radar sensor, ultrasonic sensor, thermal sensor or thermal imager.

[0035] The controller (2), for example, located at position 2, can be powered by the aircraft power system or by a separate power system. The controller integrates a power module to supply electrical energy to at least one sensor (1), for example, located at position 1. At least one sensor (1) can be integrated in various locations, for example, within the cockpit of a vehicle. Examples include within the ceiling (4), for example, within the seat (5), for example, located at position 5 (see [reference]). Figure 4 ), or within the passenger service unit (6) (e.g., located at position 6 (see Figure 4 This strategy ensures a wide field of view from all locations, as well as accurate information about the location and movement of passengers and / or crew members.

[0036] At least one sensor (1) (e.g., located at position 1) can communicate with the controller (2) (e.g., located at position 2) via wired or wireless transmission and can transmit data regarding the position and movement of passengers or crew members (7) (e.g., located at position 7). The format of the sensor data is not particularly limited and can be, for example, point cloud data, image data, and / or energy amplitude data, which can be analyzed by the processor (3) (e.g., located at position 3) of the controller (2) (e.g., located at position 2).

[0037] First, the data is preprocessed to remove all redundant information and adjust its input format (e.g., by reducing the size or scale of the data). The artificial intelligence algorithm developed to analyze data provided by at least one sensor (1) (e.g., located at position 1) integrates supervised or unsupervised methods, including deep learning algorithms (artificial neural networks), or sequentially integrates both supervised and unsupervised methods.

[0038] Previously trained AI models in similar environments categorize data into several defined classes based on the type of device to be activated. By using a large amount of training data, the AI ​​algorithm can distinguish between intentional and unintentional activation of devices and / or equipment, such as a passenger's head touching a reading light switch.

[0039] Based on the probability of different categories, the controller (2) (e.g., located at position 2) sends a trigger signal to the corresponding device and / or equipment to be activated.

[0040] The artificial model can be continuously improved during the operation and use of the system because it can be retrained using new data collected in buildings or vehicles (e.g., such as cockpits).

[0041] As an additional feature, the output of artificial intelligence algorithms (e.g., the frequency and / or probability of category activation) can be stored for subsequent design improvements and passenger and crew behavior analysis.

[0042] Figure 3 An example of data provided by at least one sensor (1) is shown in which a passenger or crew member (7) (e.g., at position 7) places his or her hand over the trash can (8) (e.g., at position 8).

[0043] Figure 4 An example of a posture recognition system in an overhead cabin compartment is shown. At least one sensor (1) (e.g., located at position 1) is located above the passenger's seat (5) and analyzes the posture of the passenger and / or crew to activate, for example, a reading light (9) (e.g., located at position 9).

[0044] Alternative embodiments This invention describes a scheme for an attitude recognition system in an aircraft, but similar applications can be used in various devices and / or equipment.

[0045] Industrial application This invention can be applied to, for example, vehicles (e.g., commercial and / or military aircraft).

[0046] This invention can be used in any area of ​​a commercial or military aircraft where devices and / or equipment can be activated by passengers and / or crew members.

[0047] List of reference numerals 1. Sensor 2 Controllers 3 processors 4. Ceiling 5 seats 6 Service Units 7 people (passengers or crew members) 8. Trash baffle 9 lights

Claims

1. A pose recognition system, comprising: At least one sensor (1). At least one controller (2), the at least one controller including at least one processor (3), the at least one controller (2) further including software, A power supply configured to supply power to the at least one sensor (1) and the at least one controller (2), At least one device and / or at least one equipment, An actuator, configured to activate the at least one device and / or the at least one apparatus, Wherein, the at least one sensor (1) is located in the area of ​​use of a static or moving environment or entity, and the at least one sensor is configured to communicate with the at least one controller (2). The number of the at least one sensor (1) is at least one less than the number of the at least one device and / or at least one equipment.

2. The posture recognition system according to claim 1, wherein, The posture recognition system is configured to communicate between the at least one device and / or the at least one equipment and the cockpit management system.

3. The posture recognition system according to claim 1 or 2, wherein, The data of the at least one device and / or the at least one equipment can be stored in the posture recognition system.

4. The posture recognition system according to any one of the preceding claims, wherein, The software in the at least one controller (2) uses an artificial intelligence algorithm, wherein the artificial intelligence algorithm is capable of integrating supervised and / or unsupervised methods, or sequentially integrating both the supervised and unsupervised methods, the supervised and / or unsupervised methods including deep learning algorithms.

5. The posture recognition system according to claim 4, wherein, The posture recognition system is configured to distinguish between intentional and unintentional activation of the at least one sensor (1) based on data analyzed by the artificial intelligence algorithm.

6. The posture recognition system according to any one of the preceding claims, wherein, Data from the at least one sensor (1) is transmitted to the at least one controller (2) via a wired connection.

7. The posture recognition system according to any one of claims 1 to 5, wherein, Data from the at least one sensor (1) is transmitted to the at least one controller (2) via wireless communication.

8. The posture recognition system according to any one of the preceding claims, wherein, The data from the at least one sensor (1) can be in the format of point cloud data, and / or image data and / or energy amplitude data.

9. The posture recognition system according to any one of the preceding claims, wherein, The at least one sensor (1) is a non-contact sensor and replaces the mechanical switch to improve hygiene and reduce maintenance.

10. The posture recognition system according to any one of the preceding claims, wherein, The at least one sensor (1) can be one or more of the following sensors (1): Optical sensors, such as optical cameras, lidar, time-of-flight sensors, or cameras; and / or radar sensors; and / or ultrasonic sensors; and / or thermal sensors; and / or thermal imagers.

11. The posture recognition system according to any one of the preceding claims, wherein, The at least one sensor (1) is located at at least one location within the area of ​​use, wherein the at least one location may be a ceiling (4), and / or a seat (5), and / or a service unit (6), and / or a toilet, and / or a kitchen, and / or a passenger and / or crew rest area.

Images