Method for determining a movement path data record of a user

Abstract

The invention relates to a method for determining a movement path data record (12) of a visitor (2) on the movement surfaces (1) of business premises, wherein a plurality of sensors (4) for radio-assisted locating of moving objects (2, 3) are arranged above the movement surfaces (1), wherein, by means of the sensors (4), locating data (7) from one or more capture points (8) of the moving objects (2, 3) are successively determined and transmitted to a computing unit (6). In order to determine the movement paths (13) of visitors (2) within business premises reliably, with little resource expenditure and by avoiding imaging methods, it is proposed that the locating data (7) are transferred from a sensor-based reference system (9) into a global reference system (10), after which the locating data (7) from different sensors (4) and successive time steps are assigned to one another to produce a movement path data record (12).

Description

[0001] Method for determining a a visitor

[0002] Technical area

[0003] The invention relates to a method for determining a movement path data set of a visitor on the movement areas of business premises, wherein a plurality of sensors for radio-assisted location of moving objects are arranged above the movement areas, wherein location data from one or more detection points of the moving objects are successively determined by the sensors and transmitted to a computing unit.

[0004] State of the art

[0005] To investigate the effectiveness of product placement in retail premises, it has already been proposed (US11615430B1) to install several cameras mounted on the ceiling. Using the images captured by these cameras and sensors for detecting signals from mobile devices, the system would be able to determine visitor location data, such as position, speed, or posture, and derive movement paths related to the visitors' purchasing decisions. The disadvantage, however, is that processing the captured images and considering the various data sources is resource-intensive, and the use of imaging cameras, in particular, raises concerns from a data protection perspective.

[0006] To analyze visitor interactions with products in retail premises, it is also known (US20220092664A1) to monitor the product sales areas with cameras and use image processing methods to detect and evaluate interactions with the products. While this is less resource-intensive and requires processing a smaller amount of data, it results in the loss of important information about users' movements within the retail premises.

[0007] Description of the invention

[0008] The invention is therefore based on the object of determining the movement paths of visitors within business premises reliably, with low resource expenditure and avoiding imaging processes, based on a method of the type described above.

[0009] The invention achieves this objective by transferring the location data from a sensor-related reference system to a global reference system, after which the location data from different sensors and successive time steps are assigned to one another to generate a motion path data set. As a result of these measures, the spatial assignment of the location data from different sensors is facilitated, the transmission of personal data between the sensors and the computing unit is avoided, and the amount of data to be transmitted is significantly reduced without significantly impairing the generation of motion path data sets. To transfer the location data from the respective sensor-related reference system to a common global reference system, the relative positions and orientations of the sensors in the global reference system can be specified.For example, a transformation matrix can be provided for each sensor to convert the location data from the sensor-specific reference system to the global reference system. It does not matter whether the transfer between the reference systems is carried out by the sensor itself before transmission or by the computing unit after transmission. To assign the location data from different sensors to one another, for example, a distance measure between the location data in the global reference system can be used, whereby those location data whose distance lies within a predetermined value range are assigned to one another. To generate a movement path data set, for example, average values ​​of the associated location data can be calculated and added to the movement path data set, so that the movement path data set comprises a series of averaged location data that depicts a visitor's movement path.The previously determined motion path data sets can be used, so that only those, for example, averaged location data are assigned to a motion path data set that lie within an expected range resulting from the location data previously assigned to the motion path data set. Location data that cannot be assigned to other location data or to a motion path data set can be discarded as interference. For example, radio sensors based on electromagnetic waves can be used, which operate in a frequency range from 3 MHz to 3000 GHz, preferably in a range from 30 MHz to 300 GHz, and particularly preferably in a range from 2 to 150 GHz.The sensors can either detect electromagnetic waves from portable devices, for example, using Bluetooth or WLAN signals, or, in a preferred embodiment, emit electromagnetic waves themselves and detect echoes reflected from moving objects. For example, radar sensors operating in a frequency range between 20 and 150 GHz, preferably around 60 GHz or around 120 GHz, can be used as sensors. To avoid mutual interference between the sensors, neighboring sensors can operate in different frequency ranges. In a preferred embodiment, the main detection direction of the sensors can be aligned perpendicular to the moving surfaces, with a conical detection area extending around the main detection direction.

[0010] To improve the correlation of positioning data and the assignment to movement path data sets, it is proposed that, for each detection point, direction, distance, and / or speed values, utilizing the Doppler effect, be transmitted as positioning data from the sensors to the processing unit. While the direction and distance data allow for absolute positioning of the moving objects in the global reference system, the speed values ​​provide information about the movement behavior of these objects, so that the correlation of positioning data to each other and to the movement path data sets can be improved, particularly in the case of spatially close objects or object overlaps and collisions. In addition, the

[0011] Speed ​​values ​​allow a more reliable determination of expected ranges based on the location data previously assigned to the respective movement path data sets. A particularly preferred method for determining speed values ​​as location data can be achieved using radar sensors.

[0012] To better distinguish location data of moving objects from detection-related interference, the signal strength at each detection point can be transmitted from the sensors to the processing unit as location data. Especially in the case of reflections or multipath propagation, the signal strength or signal-to-noise ratio can be used to distinguish the detection point of the moving object from propagation-related interference, because the signal strength of these interferences is usually significantly lower than the signal strength at the detection point of the moving object.

[0013] To improve the correlation of the location data and simultaneously avoid complex synchronization of the sensors, the invention proposes that the location data from the sensors be provided with a global timestamp. Consequently, the location data from the individual sensors can essentially be recorded independently of one another in time and subsequently assigned to predefined global logs in the computing unit, so that within a global time step, specific location data from different sensors can be assigned to each other and to the movement path data sets. The global timestamp can be continuously transmitted to the individual sensors.The interval between two transmissions can be smaller than the duration of the global time steps, so that the sensors do not require an internal timer, or the sensors have an internal timer that is adjusted to and synchronized with a common system time using the global time stamps.

[0014] To improve the reliable generation of movement path datasets beyond the detection limits of individual sensors, the location data can be assigned to a global time stamp based on their global timestamp, after which the location data for each time stamp are grouped into groups based on their values. By assigning them to global time stamps, location data from different sensors determined within the respective time step are associated with each other, so that even slightly different determination times or inaccuracies in the timers used do not affect the generation of movement path datasets. The length of the global time steps should preferably be selected so that only a correspondingly small change in the location data of those moving objects that represent visitors can be expected within a time step.By assigning the location data for each magazine to groups, multiple detection points can be combined to form an image of a moving object. This allows the moving object to be classified based on the properties of multiple location data from the group or image. In particular, this makes it possible to distinguish visitors from other moving objects, such as shopping carts. Possible methods for combining location data into groups include density-based spatial clustering of applications with noise (DBSCAN) or ordering points to identify the clustering structure (OPTICS). Combining data into groups also makes it easier to reject interference, as this is usually not statistically significant.Since visitors can only perform a certain type of real movement due to their physical conditions and the laws of nature, it is proposed that only those location data that depict a visitor's real movement be assigned to each other and / or to a movement path dataset. For example, location data whose absolute position lies outside the dimensions of a human body cannot be assigned to each other, nor can location data whose speed values ​​exceed a specified maximum speed. Furthermore, the expected range determined from a movement path dataset can be adapted according to these possible movements. For example, a Kalman filter can be used for each movement path dataset to predict the expected range, with the output values ​​of the Kalman filter being assigned to the respective movement path dataset as averaged location data.

[0015] Brief description of the invention

[0016] The drawing shows an example of the subject matter of the invention.

[0017] Fig. 1 is a schematic representation of business premises with sensors arranged above a movement area and

[0018] Fig. 2 shows a representation of selected location data over several global time steps in a global reference system.

[0019] Ways to implement the invention

[0020] Moving objects 2, 3, such as a visitor 2, are located on the movement areas 1 of business premises. The moving objects 2, 3 are detected by sensors 4 arranged above the movement areas 1 for radio-based positioning. The sensors 4 can be radar sensors, for example. Each sensor 4 is connected to a computing unit 6 via a data line 5 and sequentially determines location data 7 from one or more detection points 8 of the moving objects 2, 3. The location data 7 can include, for example, direction, distance, and / or speed values. In addition, the location data 7 can include the signal strength for each detection point as well as a global timestamp.

[0021] As can be seen in particular from Fig. 2, in a preferred embodiment the location data 7 are assigned to a global journal by the computing unit 6 on the basis of their global time stamp and transferred from a sensor-related reference system 9 into a global reference system 10.

[0022] For each global journal, the location data 7 can be combined into groups 11, for example, based on their distance from one another or based on the distance of their values ​​from one another. To generate a movement path 13 represented by a movement path data set 12, for example, average values ​​of the location data 7 of a group 11 can be calculated and added to the movement path data set 12, so that the movement path data set 12 comprises a series of averaged location data that map the movement path 13 of a visitor 2.

[0023] Starting from each motion path data set 12, an expected range 14 for location data 7 in the subsequent global journal can be determined based on the location data assigned in past global journals, for example, averaged location data. For example, a Kalman filter can be used for each motion path data set 12 to predict the expected range 14, with the output values ​​of the Kalman filter being assigned to the respective motion path data set 12 as averaged location data. Location data 7 that are due to disturbances or other moving objects 3 and lie outside the respective expected range 14 can be discarded.

Claims

Patent claims 1. Method for determining a movement path data set (12) of a visitor (2) on the movement areas (1) of business premises, wherein a plurality of sensors (4) for radio-assisted location of moving objects (2, 3) are arranged above the movement areas (1), wherein location data (7) from one or more detection points (8) of the moving objects (2, 3) are successively determined by the sensors (4) and transmitted to a computing unit (6), characterized in that the location data (7) are transferred from a sensor-related reference system (9) to a global reference system (10), after which the location data (7) of different sensors (4) and successive time steps are assigned to one another to generate a movement path data set (12).

2. Method according to claim 1, characterized in that for each detection point (8) direction, distance and / or speed values using the Doppler effect are transmitted as location data (7) from the sensors (4) to the computing unit (6).

3. Method according to one of claims 1 or 2, characterized in that the signal strength for each detection point (8) is transmitted as location data (7) from the sensors (4) to the computing unit (6).

4. Method according to one of claims 1 to 3, characterized in that the location data (7) from the sensors (4) are provided with a global time stamp.

5. Method according to claim 4, characterized in that the location data (7) are assigned to a global journal on the basis of their global time stamp, after which the location data (7) for each journal are combined into groups (11) on the basis of their values.

6. Method according to one of claims 1 to 5, characterized in that only those location data (7) are assigned to each other and / or to a Movement path data set (12) are assigned, which depict a real movement of a visitor (2).

Images