Search

EP-4739197-A1 - DEVICE AND METHOD FOR DETECTING THE BLOOD-ALCOHOL CONTENT, BAC, OF A PERSON

EP4739197A1EP 4739197 A1EP4739197 A1EP 4739197A1EP-4739197-A1

Abstract

The invention relates to a method for determining the blood-alcohol content, BAC, of a person in a vehicle (1), comprising the steps of: recording one or more images of the person in the vehicle (1) by means of one or more imaging units (2) arranged in the vehicle; analysing the images by means of a data-processing unit (3) and determining body data and facial data using a database (5); calculating, from the facial data, three-dimensional target vectors V1 and/or V2 of the mouth and/or the nose of the person with respect to an air-alcohol-measuring instrument (4) arranged in the vehicle (1), in particular an infrared laser spectrometer designed for directional and contactless measuring; activating the air-alcohol-measuring instrument (4) and measuring the breath alcohol content BRAC of the person in the region of the mouth and/or nose, taking into account the target vectors V1 and/or V2; and calculating a correction factor K taking into account the body data, in particular the sex, the weight, the age and/or the height of the person; and determining the blood-alcohol content BAC = K * BRAC.

Inventors

  • SEITNER, Florian
  • Hödlmoser, Michael

Assignees

  • emotion3D GmbH

Dates

Publication Date
20260513
Application Date
20240626

Claims (17)

  1. 1. Method for determining the blood alcohol content BAC of a person in a vehicle (1), comprising the steps a. recording, by one or more image recording units (2) arranged in the vehicle (1), one or more images of the person in the vehicle (1), b. analysis of the images by a data processing unit (3) and determination, using a database (5), of body data and facial data, c. calculation, from the facial data, of three-dimensional target vectors Vi and/or V2 from the mouth and/or nose of the person to an air alcohol measuring device (4) arranged in the vehicle (1), in particular an infrared laser spectrometer designed for directed and contactless measurement, d. activation of the air alcohol measuring device (4) and measurement of the breath alcohol content BRAC of the person in the area of the mouth and/or nose, taking into account the target vectors V1 and/or V2, e. Calculation of a correction factor K taking into account the physical data, in particular the gender, weight, age and/or height of the person, and determination of the blood alcohol content BAC = K * BRAC.
  2. 2. Method according to claim 1, characterized in that the data processing unit (3) for determining the body data from the images determines an abstracted body data model, preferably in the form of a graph, and compares the body data model with a plurality of reference models taken from the database (5).
  3. 3. Method according to claim 2, characterized in that the body data model comprises the positions of defined key points, such as wrist, arm joint, sternum, shoulder joint, knee joint, elbow joint, hip joint, or head center, as well as the distances between these key points.
  4. 4. Method according to one of claims 1 to 3, characterized in that the body data determined by the data processing unit (3) are at least the height, the body weight, the age and/or the gender of the person.
  5. 5. Method according to one of claims 1 to 4, characterized in that the data processing unit (3) for determining the facial data from the images determines an abstracted facial data model, preferably in the form of a graph, and compares the facial data model with a plurality of reference models taken from the database (5).
  6. 6. Method according to claim 5, characterized in that the facial data model comprises the positions of defined key points, such as nose, corners of the mouth, upper lip, lower lip, forehead, eyes, ears and chin, as well as the distances between these key points.
  7. 7. Method according to one of claims 1 to 6, characterized in that the facial data determined by the data processing unit (3) are at least the three-dimensional position of the nose, the mouth, the opening state of the mouth and the gaze direction vector BV of the person's head relative to a vehicle-internal coordinate system.
  8. 8. Method according to one of claims 1 to 7, characterized in that the data processing unit (3) calculates the blood alcohol content BAC taking into account a confidence value A as BAC = K * A * BRAC, wherein the confidence value A is determined taking into account the facial data, in particular taking into account the opening state of the mouth and the gaze direction vector BV of the person's head.
  9. 9. The method according to claim 8, characterized in that the data processing unit (3) determines environmental data and increases or decreases the confidence value A also taking into account the environmental data.
  10. 10. Method according to claim 9, characterized in that the environmental data are determined by querying systems located in the vehicle (1), wherein the environmental data are, for example, the driving time of the person, the current time, or the opening state of the vehicle windows.
  11. 11. Method according to claim 9 or 10, characterized in that the data processing unit (3) only activates the air alcohol measuring device (4) when a. predetermined conditions of the environmental data are met, for example a predetermined opening state of the windows of the vehicle (1) is reached, and/or b. predetermined conditions of the facial data are met, for example the opening state of the mouth or the gaze direction vector BV of the person.
  12. 12. Method according to one of claims 9 to 11, characterized in that the data processing unit (3) continuously detects whether the person consumes food or drinks by regularly activating the image recording units (2), and only activates the air alcohol measuring device (4) when a predetermined period of time has elapsed after the consumption of food or drinks, for example a period of 15 minutes.
  13. 13. Method according to one of claims 1 to 12, characterized in that the data processing unit (3) activates a warning unit in the vehicle (1) or blocks the operation of the vehicle (1) if the measured BRAC value or the calculated BAC value exceeds a predetermined value in one or, if appropriate, in several consecutive measurements.
  14. 14. Method according to one of claims 1 to 13, characterized in that a matching algorithm or a classification method from the field of machine learning is used to determine the body data and/or the facial data, in particular a neural network trained with reference images.
  15. 15. Method according to one of claims 1 to 14, characterized in that a predetermined heuristic is used to determine the body data and/or the facial data, for example the distance between the detected upper lip and lower lip of the person to determine the opening state of the mouth.
  16. 16. Computer-readable storage medium comprising instructions which cause a data processing unit (3) to carry out a method according to one of claims 1 to 15.
  17. 17. Device for determining the alcohol content of a person in a vehicle (1), comprising an image recording unit (2), a data processing unit (3), an air alcohol measuring device (4) and a database (5), characterized in that the device is set up to carry out a method according to one of claims 1 to 15.

Description

Device and method for detecting the blood alcohol content BAC of a person The invention relates to a device and a method for detecting the blood alcohol content BAC of a person. A person's level of intoxication can be measured by the alcohol concentration in the person's blood (Blood Alcohol Concentration, BAC) and by the alcohol concentration in the person's breath (Breath Alcohol Concentration, BRAC). This is particularly relevant for determining the level of intoxication of a driver. The actual impairment of a person caused by alcohol can only be reliably determined from the BAC value, but the BRAC value can be determined much more easily and, above all, non-invasively. The BAC value can be approximately calculated from the BRAC value using knowledge of various data about the person. The BRAC value can be measured using analyzers that come into direct contact with the person's mouth, so that the person's breath goes directly into the analyzer. However, there are also contactless analyzers that do not have to be placed directly next to the person's mouth. In these cases, the measurement is usually carried out using a directed infrared laser spectrometer. Such analyzers are also suitable for measuring the BRAC of a vehicle driver. However, to accurately determine the BAC value from the measured BRAC value, a large number of other data from the vehicle driver must be taken into account. When using contactless analysis devices, there is also the problem that there is no certainty as to whether the person is using the analysis device correctly and whether the conditions for a valid measurement are even met. The object of the invention is therefore to provide a method and a device for detecting the alcohol intoxication of a vehicle driver, which allows the use of a contactless analysis device for measuring the alcohol concentration in the driver's breath BRAC, and yet provides an accurate and reliable estimate of the alcohol concentration in the driver's blood BAC. These and other objects are achieved according to the invention by a method and a device according to one of the independent patent claims. A method according to the invention for determining the blood alcohol content of a person in a vehicle comprises the following steps. First, an image recording unit arranged in the vehicle takes one or more images or videos of the person in the vehicle, in particular the driver, in 2D or 3D. The image recording unit can in particular be designed as a 3D camera. A data processing unit arranged in the vehicle determines a large number of body data of the person from the image or video. The data processing unit can use an abstracted body data model, preferably in the form of a graph, to determine the body data from the images and compare the body data model with a large number of reference models taken from the database. The body data model can include the positions of defined key points, such as the wrist, arm joint, sternum, shoulder joint, knee joint, elbow joint, hip joint, or center of the head, as well as the distances between these key points. The body data determined by the data processing unit can be at least the height, body weight, age and/or gender of the person. The data processing unit also determines a large number of facial data of the person. To determine the facial data from the images, the data processing unit can determine an abstracted facial data model, preferably in the form of a graph, and compare the facial data model with a large number of reference models taken from the database. The facial data model can include the positions of defined key points, such as the nose, corner of the mouth, upper lip, lower lip, forehead, eyes, ears and chin, as well as the distances between these key points. The facial data determined by the data processing unit can be at least the three-dimensional position of the nose, the mouth, the opening state of the mouth and the gaze direction vector BV of the person's head relative to a coordinate system inside the vehicle, in particular relative to an air alcohol measuring device arranged in the vehicle. It can also be detected whether the person is wearing a mask covering the mouth and/or nose. To determine the body data and/or the facial data, a matching algorithm or a classification method from the field of machine learning can be used, in particular a neural network trained with reference images. However, predetermined heuristics can also be used, for example the distance between the detected upper lip and lower lip of the person to determine the opening state of the mouth. From the facial data, the data processing unit calculates three-dimensional target vectors Vi and/or V2 of the person's mouth and/or nose relative to an air alcohol measuring device arranged in the vehicle, in particular an infrared laser spectrometer designed for directed and contactless measurement. The target vectors V1 and V2 indicate the three-dimensional distance and angle between t