DE112021004191T5 - Procedure for checking the presence of objects in an unauthorized location of an aircraft cabin - Google Patents

Abstract

Method for checking the presence of objects in an illegal location of an aircraft cabin, comprising the following steps:detecting (100) an object placed in the illegal location using pressure values measured on the surface of the illegal location;determining (200 ) the shape of the detected object based on the pressure values obtained;comparing (300), by an algorithm trained by machine learning, the shape of the detected object with reference shapes associated with a footprint or shoe print and stored in a database; and the algorithm identifying (400) a shape that is not associated with a footprint or shoeprint.

Description

technical field

The present invention relates generally to measures for the safety of passengers on board an aircraft and in particular to checking the proper organization and stowage of passenger luggage in an aircraft cabin.

More precisely, the invention relates to a control method for detecting the presence of objects in an unauthorized location of an aircraft cabin.

The invention also relates to a system for checking the presence of objects in an unauthorized location in an aircraft cabin.

During the take-off, landing and taxiing phase of an aircraft, the flight crew must check whether the luggage and other personal belongings of the passengers are stored in appropriate permitted places in the cabin.

Such a verification procedure requires the inspection of all unauthorized places in the cabin, especially at the feet of the passengers, between the seats in a row and at the emergency exits, which are not allowed to be served.

In fact, the presence of objects in unauthorized locations poses risks in the event of severe turbulence while taxiing. Also, it is very important that escape routes and areas near emergency exits are not blocked in an emergency.

However, the checks carried out by the on-board staff are selective and do not make it possible to continuously guarantee the safety of the passengers

In addition, this control is long and tedious.

State of the art

Solutions aim to carry out this control using object detection cameras, or using distance sensors or temperature sensors combined with an object detection algorithm based on artificial intelligence, machine learning or automatic learning technology. , based.

However, these techniques require numerous learning attempts in order to be able to cover and recognize all cases of luggage and relative positions between luggage and their combinations with a passenger's legs and feet.

Disclosure of Invention

The object of the invention is therefore to remedy these disadvantages and to propose a simple and reliable method for detecting objects in unauthorized locations in an aircraft cabin, the aim of which is to shorten the time it takes to check for the presence of these objects.

• die Detektion eines an dem unerlaubten Ort abgelegten Gegenstands anhand von Druckwerten, die an der Oberfläche des unerlaubten Ortes gemessen werden;
• die Bestimmung der Form des detektierten Gegenstands anhand der erhaltenen Druckwerte;
• den Vergleich, durch einen durch maschinelles Lernen trainierten Algorithmus, der Form des detektierten Gegenstands mit Referenzformen, die einem Fuß- oder Schuhabdruck zugeordnet und in einer Datenbank gespeichert sind; und
• die Identifikation einer Form, die nicht einem Fuß- oder Schuhabdruck zugeordnet ist, durch den Algorithmus.

A method for checking the presence of objects in an unauthorized location of an aircraft cabin is therefore proposed, which comprises the following steps:

• detecting an object placed in the illegal location from pressure values measured at the surface of the illegal location;
• determining the shape of the detected object from the pressure values obtained;
• comparing, by an algorithm trained by machine learning, the shape of the detected object with reference shapes associated with a footprint or shoeprint and stored in a database; and
• the identification by the algorithm of a shape that is not associated with a footprint or shoeprint.

Preferably, the control method includes reporting the identification of the item with a shape not associated with a footprint or shoeprint at the unauthorized location.

Preferably, the notification of the identification of the object having a shape other than a footprint or shoeprint at the unauthorized place is made by notification on a control screen or by an indicator light.

Advantageously, the pressure values can be obtained using one or more pressure sensors selected from a piezoelectric sensor, a capacitive sensor and a resistive sensor.

Advantageously, the control method may comprise measuring the dimensions of the shape of the detected object and comparing them with reference dimensions associated with a footprint or shoeprint.

According to one embodiment, the control method can include a step of detecting a movement of the detected object.

According to another embodiment, the control method may comprise measuring the force per unit area exerted on the illegal location and comparing it to reference values of the force per unit area exerted by a foot or a shoe.

• Mittel zur Detektion eines Gegenstands, die mehrere Drucksensoren umfassen, die geeignet sind, den auf die Oberfläche des unerlaubten Ortes ausgeübten Druck zu messen;
• Mittel zur Bestimmung der Form des detektierten Gegenstands anhand der erhaltenen Druckwerte;
• eine Datenbank, welche Referenzformen enthält, die einem Fuß- oder Schuhabdruck zugeordnet sind;
• Mittel zur Ausführung eines durch maschinelles Lernen trainierten Algorithmus, der dafür ausgelegt ist, die Form des detektierten Gegenstands mit in der Datenbank gespeicherten Referenzformen zu vergleichen, und dafür ausgelegt ist, eine Form zu identifizieren, die nicht einem Fuß- oder Schuhabdruck zugeordnet ist.

The invention also relates to a system for checking the presence of objects in an unauthorized location in an aircraft cabin, comprising:

• means for detecting an object, comprising a plurality of pressure sensors capable of measuring the pressure exerted on the surface of the illegal site;
• means for determining the shape of the detected object from the pressure values obtained;
• a database containing reference shapes associated with a footprint or shoeprint;
• Means for executing a machine learning trained algorithm adapted to compare the shape of the detected object to reference shapes stored in the database and adapted to identify a shape that is not associated with a footprint or shoe print.

Preferably, the control system comprises a device for reporting the detection of the object with a shape that is not associated with a footprint or shoeprint at the unauthorized location.

Preferably, the multiple pressure sensors include one or more pressure sensors selected from a piezoelectric sensor, a capacitive sensor, and a resistive sensor.

character list

• [1] ein System zur Kontrolle des Vorhandenseins von Gegenständen an einem unerlaubten Ort einer Luftfahrzeugkabine gemäß einer Ausführungsform der Erfindung veranschaulicht,
• [2] ein Flussdiagramm ist, welches ein Verfahren zur Kontrolle des Vorhandenseins von Gegenständen an einem unerlaubten Ort einer Luftfahrzeugkabine gemäß einer Ausführungsform der Erfindung veranschaulicht.

Other objects, advantages and features will appear from the following description, given for illustrative purposes only and with reference to the attached drawing, in which:

• [ 1 ] illustrates a system for checking the presence of objects in an unauthorized location of an aircraft cabin according to an embodiment of the invention,
• [ 2 ] is a flow diagram illustrating a method for checking the presence of objects in an unauthorized location of an aircraft cabin according to an embodiment of the invention.

Detailed description of an embodiment

1 Figure 1 illustrates a system for checking 1 the presence of objects in an unauthorized location of an aircraft cabin.

In the example shown, a prohibited location is understood to mean a location on the floor of the cabin of the aircraft, at the feet of the passengers between the seats in a row of seats or at the feet of the passengers at an emergency exit.

The control system 1 comprises detection means 2 which make it possible to detect an object located in an unauthorized location.

The detection means 2 comprise several pressure sensors capable of measuring the pressure exerted on the surface of the illegal site, enabling the presence of an object to be detected.

Preferably, the pressure sensors are placed on the entire floor area between the seats so that they cover all unauthorized places of the aircraft cabin.

Determination means 3 are also provided, which are suitable for determining the shape of the detected object from the pressure values obtained from the detection means 2 .

Reference shapes that are associated with a footprint or a shoe print are combined in a database 4 .

A footprint or shoe print means a print corresponding to one or two feet and/or one or two shoes belonging to a passenger.

The control system 1 also includes means for executing 5 an algorithm trained by machine learning.

Machine learning, translation of the English term “machine learning”, means an automatic learning method based on artificial intelligence technology for automatically performing a step of comparing parameters through a previously trained algorithm.

The algorithm executed by the execution means 5, preferably a computing device, is designed to compare the shape of the detected object with reference shapes stored in the database 4.

The algorithm is also designed to identify a shape that is not associated with a footprint or shoeprint.

The invention also relates to a method for checking the presence of objects in an unauthorized location of an aircraft cabin, an embodiment of which is shown in 2 is illustrated.

The control method includes a first step 100 of detecting an object placed in an unauthorized location. The detection step 100 is carried out on the basis of pressure values measured on the surface of the illegal site in order to assess the bearing pressure of any object present on the ground.

The pressure values are obtained using one or more pressure sensors.

In this regard, the control system 1 may include one or more piezoelectric, capacitive, or resistive sensors.

For example, the pressure sensors can either be glued to the structure of the floor of the aircraft cabin and covered by the carpet covering the floor, or glued to the underside of the carpet, or integrated into the structure of the carpet's lower support.

A second step 200 of the control method consists in determining the shape of the detected object from the pressure values measured in the illegal location and their distribution on the surface of the specific location.

This determination of the shape of the detected object thus corresponds to a reconstruction of the geometric contour of the detected object using the measured pressure values and their distribution at the unauthorized location.

When the shape of the detected object has been determined, the method continues with a step 300, performed by a machine learning algorithm, of comparing the shape of the detected object with reference shapes or curves stored in the database 4 and a footprint or correspond to a shoeprint.

These reference shapes or curves are created, for example, from anthropometric data relating to shoe or foot sizes and comprising a length, a width at the arch area, a width at the heel area; or from biometric data including a minimum and a maximum pressure value and pressure distribution curves along the length and width of a considered shod or bare foot obtained from measurements of the pressure of shod or bare feet on a supporting surface. Such biometric data are data measured in the usual way during biomechanical studies related to postural or other assessments and allow to define reference shapes or curves that are stored in the database 4.

The algorithm compares the shape of the detected object to the reference shapes or curves to recognize a footprint or shoeprint described by that shape.

The algorithm is previously trained and validated using conventional prior learning techniques available to those skilled in the art on shapes of objects that may be present in reality at the illicit location, including footprints, shoeprints, shapes of luggage and other personal items of passengers and others to classify the shape of a detected object as a footprint or shoeprint allowed in the area or another object not allowed in the area.

Based on this comparison, the algorithm performs in a later step 400 the identification of a shape that is not associated with a footprint or shoe print.

The screening method then enables automatic detection of an object other than a passenger's foot or shoe without the need to detect different shapes of luggage and other passenger personal effects that may be present at the unauthorized location.

If a shape not associated with a footprint or shoeprint is detected, evidence of the presence of an illegal object at the illegal location, the control method may include a step 500 of reporting the identification of the object by notification in order to warn the crew on board.

To this end, the control system 1 can comprise a warning device 6, for example having a control screen and/or an indicator light.

In order to improve the detection of an object in an illegal location, the control method 1 can also include measuring the dimensions of the shape of the object detected and comparing these dimensions with reference dimensions associated with a footprint or shoeprint. For example, reference dimensions corresponding to a foot or two, or a large size shoe or two, facilitate the classification of any detected item whose shape exceeds those reference dimensions as an illegal item.

It is also possible to further provide a step of detecting a movement of the detected object, so as to be able to distinguish, when a shape is detected in an illegal place, between the presence of a stationary object and the presence of a moving foot or shoe.

This detection of a movement is performed based on a comparison of the pressure measurement data obtained in the prohibited area at different points in time in order to detect changes in these measurement values over time.

If the pressure exerted by the detected object changes during a predetermined period of time, the moving object is considered to be a foot or shoe in motion.

It is also possible to foresee a step of measuring the force exerted per unit area on the illegal site and comparing it with reference values of the force exerted per unit area by a foot or a shoe in order to more precisely between to distinguish an object associated with a passenger's foot or shoe and a prohibited object.

Claims (10)

A method of checking for the presence of objects in an unauthorized location of an aircraft cabin, comprising the steps of: detecting (100) an object placed in the illegal location from pressure values measured at the surface of the illegal location; determining (200) the shape of the detected object from the pressure values obtained; comparing (300), by a machine learning-trained algorithm, the shape of the detected object with reference shapes associated with a footprint or shoeprint and stored in a database; and identifying (400) by the algorithm a shape that is not associated with a footprint or shoeprint. control procedures claim 1 , characterized in that it comprises reporting (500) the identification of the object with a shape that is not associated with a footprint or shoe print at the prohibited location. control procedures claim 1 or 2 characterized in that the notification (500) of the identification of the object with a shape not associated with a footprint or shoeprint at the unauthorized place is made by notification on a control screen or by an indicator light. Control method according to any one of the preceding claims, characterized in that the pressure values are obtained using one or more pressure sensors chosen from a piezoelectric sensor, a capacitive sensor and a resistive sensor. Control method according to any one of the preceding claims, characterized in that it comprises measuring the dimensions of the shape of the detected object and comparing them with reference dimensions associated with a footprint or shoeprint. Control method according to any one of the preceding claims, characterized in that it comprises a step of detecting a movement of the detected object. Control method according to any one of the preceding claims, characterized in that it comprises measuring the force per unit area exerted on the illegal site and comparing it with reference values of the force per unit area exerted by a foot or a shoe. System for checking the presence of objects in an illegal location in an aircraft cabin, comprising: object detection means (2) comprising a plurality of pressure sensors capable of measuring the pressure exerted on the surface of the illegal location; Means for determining (3) the shape of the detected object based on the pressure values obtained; a database (4) containing reference shapes associated with a footprint or shoeprint; Means for executing (5) an algorithm trained by machine learning, adapted to compare the shape of the detected object with reference shapes stored in the database (4) and adapted to identify a shape that does not correspond to a foot or shoeprint is assigned. control system after claim 8 , characterized in that it comprises a device for reporting (6) the detection of the object with a shape not associated with a footprint or shoeprint in the unauthorized place. control system after claim 8 or 9 , characterized in that the plurality of pressure sensors comprises one or more pressure sensors selected from a piezoelectric sensor, a capacitive sensor and a resistive sensor.

Images