COMPUTER-ASSISTED METHOD FOR ANALYSING PERSON-RELATED ACCIDENT DATA

20210319238 · 2021-10-14

    Inventors

    Cpc classification

    International classification

    Abstract

    A computer-assisted method for analyzing person-related accident data of one or more vehicle occupants, the person-related accident data including at least image data of a video sequence of the vehicle occupants during an accident. The computer-assisted method includes at least one first step of detecting a pattern on the basis of the image data, and at least one second step of comparing the detected pattern with a number of previously stored patterns.

    Claims

    1-14. (canceled)

    15. A computer-assisted method for analyzing person-related accident data of one or more vehicle occupants, wherein the person-related accident data comprise at least image data of a video sequence of the vehicle occupant during an accident, the computer-assisted method comprising: one first step of detecting a pattern using the image data; and one second step of comparing the detected pattern with a number of previously-stored patterns, wherein this detected pattern comprises a movement sequence of the vehicle occupant during the accident, wherein a detected and/or previously-stored pattern comprises a time sequence of acceleration processes of a head and/or a neck and/or a thorax and/or a torso and/or extremities of the vehicle occupant and/or a time sequence of displacements of a head and/or a neck and/or a torso and/or a thorax and/or extremities of the vehicle occupant, wherein the previously-stored patterns are respectively assigned medical diagnostic data and that the method further comprises a third step of outputting medical diagnostic data that are assigned to the detected pattern.

    16. The computer-assisted method according to claim 15, wherein the medical diagnostic data comprise at least one medical diagnosis and/or one piece of information about medical long-term consequences.

    17. The computer-assisted method according to claim 15, wherein the medical diagnostic data comprise an indicator value specifying the degree of a possible severity of an injury of the vehicle occupant.

    18. The computer-assisted method according to claim 15, wherein the person-related accident data further comprise physical operating data of the vehicle immediately before and/or during the accident.

    19. The computer-assisted method according to claim 15, which is conducted by using an algorithm for implementing machine learning.

    20. The computer-assisted method according to claim 19, wherein the algorithm is suitable for detecting a number of patterns from image data in advance using a set of personal accident data provided, which image data are saved as previously-stored patterns.

    21. The computer-assisted method according to claim 15, wherein the previously-stored patterns are saved in an accident database.

    22. A device for computer-assisted analyzing of person-related accident data of one or more vehicle occupants, wherein the person-related accident data comprise at least image data of a video sequence of the vehicle occupant during an accident, the device comprising: a detection unit for detecting a pattern using the image data; and a comparison unit for comparing the detected pattern, comprising a movement sequence of the vehicle occupant during the accident with a number of previously-stored patterns, wherein the detected and/or previously-stored patterns comprise a time sequence of acceleration processes of a head and/or a neck and/or a thorax and/or a torso and/or extremities of the vehicle occupant and/or a time sequence of displacements of a head and/or a neck and/or a torso and/or a thorax and/or extremities of the vehicle occupant, wherein the previously-stored patterns respectively are assigned medical diagnostic data and the device furthermore comprises and output unit for medical diagnostic data, assigned to the detected pattern.

    23. An accident assistance system comprising a device according to claim 22 and a high-speed camera for capturing image data of a video sequence of the vehicle occupant during an accident and a data memory for saving the image data recorded by the high-speed camera.

    24. The accident assistance system according to claim 23, further comprising an operating data recording unit for recording physical operating data of a vehicle.

    Description

    [0028] Further properties and purposes of the invention ensue from the description of the illustrative examples using the appended drawings.

    [0029] FIG. 1 is a schematic view of a vehicle with a system for capturing person-related accident data connected to a device for computer-assisted analysing of person-related accident data,

    [0030] FIG. 2 is a schematic view of a device for computer-assisted analysing of person-related accident data according to an illustrative example of the present invention and

    [0031] FIG. 3 shows schematically a method for analysing person-related accident data according to the present invention.

    [0032] FIG. 1 shows a vehicle 1 with a vehicle occupant 3 and a schematically-shown vehicle-internal system 2 for capturing person-related accident data connected via a data connection to a device 20 according to the invention for computer-assisted analysing of person-related accident data. The vehicle 1 shown in FIG. 1 as an example is a motor vehicle, but it may, for example be designed as a utility vehicle.

    [0033] The vehicle-internal system 2 for recording person-related accident data comprises at least one camera designed as a high-speed camera 4 for capturing image data, a data memory not shown in more detail designed as a circular memory and a transmission unit not shown in more detail for transmitting the data stored in the circular memory to the device 20. Optionally, the system 2 also comprises an operating data capture unit (not shown) for capturing physical operating data of the vehicle 1.

    [0034] The high-speed camera 4 of the system 2 is designed and arranged in the vehicle 1 in such a way that it is suitable for capturing a region of the vehicle interior, in which the vehicle occupant 3 is located, i.e. to record a video sequence in the form of images of the vehicle occupant 3. In so doing, the high-speed camera 4 does not have to capture the entire body of the vehicle occupant 3, rather it is sufficient to capture an upper section of their body, particularly the head 5, neck and upper body and/or shoulder region. Preferably, the camera also captures distal regions of the body.

    [0035] The high-speed camera 4 exhibits a high recording speed, for example, of at least 700 fps, preferably at least 1000 fps. The images of the video sequence are recorded by the high-speed camera 4 as digital image data.

    [0036] The high-speed camera 4 is connected by a data connection, for example, a data cable, to the circular memory. The circular memory is suitable for storing the image data 20 recorded by the high-speed camera 4 in digital form. Alternatively, the circular memory can also be an internal memory of the high-speed camera 4.

    [0037] A circular memory is a digital memory with specified size or even storage capacity in which data, in the present case, including image data captured by the high-speed camera, can be continuously stored. If the maximum size of the circular memory is reached and the circular memory is full, then the respective oldest saved element is overwritten, so that the data is saved in what is called a loop. Therefore, the graphical representation of the memory is a ring shape. The circular memory may, for example, be implemented with a suitable piece of software, by means of which the storage and reading of data in a digital memory is controlled accordingly. Preferably, the size of the circular memory or its storage capacity is sufficient, during a time interval of several seconds, for example 10 s or 20 s, to store the images recorded by the high-speed camera and any physical operating data of the vehicle to be saved, before these are overwritten again.

    [0038] The optional operating data capture unit is preferably also connected by a data connection, for example, a data cable, to the circular memory, for saving the physical operating data of the vehicle 1, recorded by the operating data capture unit, in the circular memory. The operating data capture unit may, for example, be designed as speed sensor and/or as a position sensor, for example, as GPS and/or as an accelerometer for capturing an instantaneous speed or even position or even acceleration of the vehicle 1 as operating data.

    [0039] Furthermore, the circular memory is connected by a data connection, for example, a data cable, to the transmission unit of the system 2. The transmission unit is suitable for transmitting the image saved in the circular memory and optionally the operating data as person-related accident data 10 by means of a wireless transmission method to the device 20, preferably encrypted. To do this, the transmission element exhibits a data interface not shown.

    [0040] The system 2 may comprise even further components not described in more detail such as, for example, a control unit for controlling the individual components of the system 2, a sensor for establishing an accident, an illumination device for illuminating the vehicle interior, an analysis unit for analysing data and/or other data memories.

    [0041] When the system 2 is operating, the high-speed camera 4 continuously films the vehicle occupant 3 while the vehicle 1 is travelling, i.e. it records image data of the vehicle occupant 3, and transmits these to the circular memory. Optionally, at the same time by the operating data capture unit, preferably continuously, operating data of the vehicle 1 are recorded and also transmitted to the circular memory. As long as no accident occurs, the data stored in the circular memory are respectively overwritten by new data again depending on a certain recording time that depends, among other things, on the recording capacity of the circular memory.

    [0042] During an accident, this continuous data storage is interrupted, so that the data present in the circular memory is not overwritten any more. Therefore, after the end of data storage, image data and optionally physical operating data in the circular memory are present for a period in which the accident occurred. These are then transmitted as person-related accident data 10 by the transmission unit to the device 20 for analysing person-related accident data. Alternatively, the person-related accident data may also be transmitted to an external location, for example, to a rescue control centre and/or a further medical facility and/or a data memory external to the vehicle, for example, an external server or a cloud and be provided by this external location for usage in the device 20. The transmission of the person-related accident data 10 thereby occurs wirelessly, for example, by radio, particularly mobile radio or Internet, and may, for example, be done together with an electronically-made emergency call (known as an eCall). Preferably the person-related accident data 10 are transmitted in encrypted format, for example, by means of OpenPGP, wherein the person-related accident data 10 are encrypted by the transmission unit by means of a public key generated by the device 20 or even the external location and only decryptable again by the corresponding private key of the device 20 or even the external location.

    [0043] Alternatively, the device 20 for computer-assisted analysing of person-related accident data may be implemented in a mobile device (not shown in the figures), which guides a first aider and, for example, may be connected by a data connection, for example, a data cable, to the transmission unit. The mobile device together with the vehicle-internal system 2 is an example of an accident assistance system. This is used for timely recognition of possible injuries of the vehicle occupants.

    [0044] The device 20 shown schematically in FIG. 2 for computer-assisted analysing of person-related accident data comprises at least one detection unit 21, a comparison unit 22 and an output unit 23. The operation of the device 20 for computer-assisted analysing of person-related accident data is described in the following with reference to FIG. 3.

    [0045] In a first step S1 (see FIG. 3), of a computer-assisted method according to the invention for analysing person-related accident data, the detection unit 21 detects a pattern M(x) from the image data of the person-related accident data 10. Optionally, the detection unit may use, in addition to the image data, the physical operating data of the vehicle 1 for detecting the pattern M(x).

    [0046] A pattern M generally describes one or more movement sequences of the vehicle occupant 3 during the accident. For example, a pattern may comprise an acceleration or even a time sequence of accelerations of at least one section of the body of the vehicle occupant 3. Alternatively or in addition, a pattern may comprise a time sequence of relative and/or absolute displacements of at least one section of the body of the vehicle occupant 3. Such a section of the body may, for example, be the head 5 and/or the neck and/or the thorax and/or the torso and/or an extremity, e.g. an arm of the vehicle occupant 3. Particularly, in so doing, relative displacements and/or accelerations of various body parts in relation to each other may be considered such as, for example, a relative displacement of the head 5 in relation to the shoulder region or upper body of the vehicle occupant 3. Optionally, a pattern may also comprise the physical operating data, for example, the acceleration and/or speed, of the vehicle immediately before and/or during the accident.

    [0047] Subsequently, the comparison unit 22 compares, in a second step S2 the pattern M(x) detected by the detection unit with a number of previously-stored patterns M(1) to M(n). The previously-stored patterns may, in this case, for example, be stored in a database wherein the database may be an internal memory of the device 20, or the comparison unit 22 accesses a device-external database (also designated as an accident database). A result of the comparison may be, in this case, that the detected pattern M(x) corresponds to one of the previously-stored patterns M(1) to M(n), or that the recognised pattern M(x) does not correspond to any of the previously-stored patterns.

    [0048] For example, the database may comprise a number of n patterns M(1) to M(n), wherein n is an natural number greater than or equal to 1 (n≥1). Preferably, the database comprises several patterns M(1) to M(n), i.e. more than one pattern, therefore n>1. For patterns M(1) to M(n), this preferably relates to various patterns.

    [0049] Preferably, further medical diagnostic data D(1) to D(p) are saved (or even sets of medical diagnostic data), where p is also a natural number greater than or equal to 1 (p≥1). Preferably, the number p corresponds to the medical diagnostic data D(1) to D(p) of the number n of patterns previously stored in the database (i.e. p=n), wherein each of the n patterns M(1) to M(n) are precisely assigned one set of medical diagnostic data D(1) to D(p).

    [0050] A set of medical diagnostic data D(i) (1≤i≤p) comprises at least one medical diagnosis and/or a piece of information about medical long-term consequences. The medical diagnostic data, i.e. the medical diagnosis and/or the information about medical long-term consequences in this case relates to information obtained from a previous accident of a particular other vehicle occupant, in which the pattern M(i) has been detected, to which the set of medical diagnostic data D(i) is assigned.

    [0051] In this case, the medical diagnostic data comprise information on the type and/or severity of injuries of the vehicle occupants. In so doing, not only are injuries established immediately after the accident considered, but also effects occurring only later on, i.e. days, weeks or even months after the accident (long-term consequences).

    [0052] Alternatively or in addition, a severity of the injuries may be assigned to a set of medical diagnostic data D(i). To do this, the injuries may be classified, for example, as slight, moderate, serious, very serious or life-threatening and a corresponding indicator value (e.g. “slight” or “moderate” or “serious” or “very serious” or “life-threatening” or according to a scale of 1 (slight) to 5 (life-threatening)) may be stored in the medical diagnostic data D(i).

    [0053] In the event that in step S2, the comparison of the detected pattern M(x) with the previously-stored patterns M(1) to M(n) results in that the detected patterns M(x) corresponding to one of the previously-stored patterns, i.e. M(x)=M(j) with 1≤j≤n, the output unit 23 (see FIG. 2) outputs in step S3 (see FIG. 3) the diagnostic data D(j) corresponding to this pattern and/or the pattern M(j).

    [0054] For saving the previously-stored pattern and the relevant medical diagnostic data in a database, for example, it is possible to proceed as follows: respectively pairs of person-related accident data and the medical diagnostic data assigned to this are provided. The person-related accident data and medical diagnostic data may, for example, be determined from previous accidents. By means of a suitable algorithm, from the image data, possibly in connection with the physical operating data of the vehicle and/or the medical diagnostic data, corresponding patterns are detected. Preferably, the algorithm is based on a method of machine learning, i.e. it is suitable for capturing corresponding patterns from the data independently and only using the data provided.

    [0055] Each pattern thus detected is then saved as a previously-stored pattern with the corresponding medical diagnostic data in the database or in another data memory.

    [0056] In the event that in step S2 of the comparison of the detected pattern M(x) with the previously-stored patterns M(1) to M(n) it results that the detected pattern M(x) does not correspond to any of the previously-stored patterns, the output unit is preferably designed to output a corresponding piece of information, for example, in the form of a text and/or a visual or audible signal. Alternatively or in addition, the output unit may also output the detected pattern M(x) itself.

    [0057] Optionally, the detected pattern M(x), which corresponds to none of the previously-stored patterns M(1) to M(n), is then also stored i.e. saved in the database, and in the presence of corresponding diagnostic data that, for example, comprise an initial and/or follow-on diagnosis and/or information about long-term consequences, these may then be assigned later on to the newly-stored pattern M(x) and be saved in the database in connection with this.