Luminaire system with camera-based sensors for object detection and tracking

Abstract

The invention relates to a method for setting up a camera-based sensor of a luminaire, comprising the following: commissioning the camera-based sensor at the location of the luminaire, recording positions and movements of objects by means of the camera-based sensor in a detection range of the camera-based sensor, analyzing the recorded object positions and object movements with respect to a frequency of position changes of the objects, and determining an activity level of the objects from the frequency of position changes, recording the position and activity level of each object, determining active and inactive movement areas from the object positions and the activity levels, wherein areas in which object positions of objects with an activity level above a predetermined limit were recorded are defined as active movement areas,and the remaining areas of the detection range of the camera-based sensor are defined as non-active motion areas, and limiting the motion detection by the camera-based sensor system during operation of the luminaire to the active motion areas. Furthermore, the invention relates to a computer program and a control device for a luminaire with a camera-based sensor.

Description

[0001] The invention relates to a method for setting up a camera-based sensor of a luminaire as well as a computer method therefor and a control device for a luminaire with a camera-based sensor.

[0002] For indoor and outdoor luminaires, especially for industrial luminaires and traffic lighting systems, sensor systems for detecting objects and controlling the luminaire are essential. In addition to standard systems such as PIR, radar, ultrasonic, or infrared sensors, camera systems that capture a visible image are increasingly being used. However, for all of these systems, the detection range of the sensor system must be adjusted during luminaire installation. This can be a complex undertaking.

[0003] One option for setting up luminaires with camera-based sensor systems is to transmit the image data from the sensor to an external location in order to use the image data to externally adjust the detection range of the camera-based sensor system used to operate the luminaire. The goal is to exclude areas with interference or unwanted movements, which should not cause a change in the lighting level caused by the luminaire, when setting up the sensor's detection range. This is possible in principle by evaluating the sensor image at an external location and defining the detection range from there. However, this is a very complex process and requires high-bandwidth data transmission from the location of the camera-based sensor system to the external location.

[0004] Systems are already known in the prior art that are intended to simplify the setup of the camera-based sensor system at the luminaire without image transmission to an external location. For example, European patent application EP 4 102 939 A1 describes a method for parameterizing a luminaire that supports guided commissioning of the luminaire with a camera-based sensor system. The detection range of the camera-based sensor is supported by a graphical user interface at the luminaire location. However, commissioning the luminaire with a camera-based sensor system still requires manual settings via the graphical user interface at the luminaire location, which the user must make.

[0005] The aim of the present invention is to further simplify the commissioning of a luminaire with a camera-based sensor, in particular to eliminate the need for manual adjustments by a user.

[0006] The object is achieved by a method according to claim 1 and by a computer program according to claim 8 and a control device according to claim 9.

[0007] The method of the present invention is carried out fully automatically by a control device of a luminaire with a camera-based sensor at the luminaire location during setup of the camera-based sensor system. After commissioning of the camera-based sensor at the luminaire location, the positions and movements of objects are recorded using the camera-based sensor within a detection range of the camera-based sensor. This detection range still encompasses the maximum possible detection range of the camera-based sensor. The aim of the setup is to limit this detection range to an area used during operation of the luminaire. To do this, the area in which moving objects are to be detected, which are to trigger a change in the luminaire's illumination level, must be determined.For example, traffic areas in which road users (vehicles or people) are moving are monitored, while adjacent areas in which, for example, only vegetation predominates are not monitored, so that plants moving due to the influence of wind do not trigger a change in the lighting level of the light. It is therefore expedient to limit the maximum possible detection range of the camera-based sensor to partial areas (in the previous example, traffic routes). For this purpose, the method according to the invention provides for the recording of positions and movements in the entire detection range of the camera-based sensor after the initial commissioning of the camera-based sensor at the location of the light, the recorded object positions and object movements are analyzed with regard to the frequency of position changes of the objects and an activity level of the objects is determined from the frequency of position changes.The position and activity level of each object are recorded. This recording is also referred to as creating a heatmap. From these recordings, active and inactive motion zones can be defined. Areas in which objects with frequent position changes are detected are defined as active motion zones, while areas with no or rarely moving objects are defined as inactive motion zones. To define active and inactive motion zones, the object positions of objects with an activity level above a specified threshold (which can vary depending on the type of luminaire) are recorded and defined as active motion zones. The remaining areas of the maximum detection range of the camera-based sensor system are defined as inactive motion zones.This recording and analysis of object positions and activity levels is carried out for a predetermined period of time, e.g. over several hours or several days, over many iterations. At the end of the process, the camera-based sensor system is set so that it only detects movements in the active movement area. The remaining inactive movement areas are hidden from motion detection so that during subsequent regular operation of the luminaire, movements are only detected in the active movement areas. This means that the luminaire is set automatically in the camera-based sensor system, i.e. without manual access to the control system. In principle, it is possible to restart this process by resetting the luminaire, e.g. if the luminaire with the camera-based sensor system is moved to another location or if the area to be illuminated, e.g. the traffic area, is redesigned.The previous example referred to traffic routes, e.g., in public spaces. However, the method is equally applicable indoors. For example, for office lighting, it is often desired that the camera-based sensor only detects those areas where people move, e.g., along corridors, while adjacent areas are not to be detected. The principle for setting up the luminaire is the same as previously described for traffic routes.

[0008] According to a preferred embodiment, the step of analyzing the recorded object positions and movements with respect to the frequency of position changes of the objects is repeated at intervals of at least one minute or after recording at least 1,000 data sets concerning the position and movement of objects, and the step of recording the position and activity level of each object is adjusted according to the repeated analysis. The plurality of data sets allows for a statistically reliable determination of the frequency of position changes of objects that are typical for the area to be illuminated.

[0009] According to a preferred embodiment, during the repeated analysis of the recorded object positions and movements, an automatic plausibility check is additionally performed, which determines position jumps of the object based on the recorded movement data of the object. The method further comprises: deleting recorded data records that contain implausible movement data of an object, wherein implausible movement data is movement data in which position jumps above a predetermined threshold have been detected, and updating the recordings of the positions and the activity level of each object only taking into account the non-deleted data records. The plausibility check excludes measurement artifacts. A position jump cannot occur for a normally moving object. It is therefore highly likely that this is a measurement error.These erroneous data records are eliminated through plausibility checks.

[0010] According to a preferred embodiment, the deletion of data records containing implausible movement data of an object occurs after at least 24 hours or after at least 50,000 data records concerning the position and movement of objects have been recorded using the camera-based sensor. The deletion of implausible movement data occurs only at longer intervals to obtain a more reliable statistical result.

[0011] According to a preferred embodiment, the step of determining active and inactive movement ranges from the object positions and activity levels occurs after at least 48 hours or after at least 100,000 recorded data sets concerning the position and movement of objects using the camera-based sensor. The determination of the active and inactive movement ranges only occurs after a large number of iterations to improve the reliability of the result. In particular, in conjunction with the plausibility analysis, erroneous data is eliminated, or individual incorrect measurements are averaged out using statistics.

[0012] According to a preferred embodiment, the step of recording positions and movements of objects using the camera-based sensor further comprises: recording light values in the detection area using the camera-based sensor; and after at least one minute or after at least 1,000 recorded data sets of the light values, analyzing changes in the light values and comparing them with recorded object positions. The light values can be used to assign a weighting value to the recorded data set. If little light is measured, the weighting is low, so that this data set is given a correspondingly lower weighting in the subsequent evaluation.

[0013] According to a preferred embodiment, the method further comprises providing a user application configured to allow a user to enter limit values that are used in the analysis of the recorded object positions and movements in order to discard data records that exceed the defined limit values. Although the user application requires manual settings by the user, the presetting can reduce the time required to set up the luminaire using the method according to the invention and / or improve the result of the determination because a rough preselection for the active movement areas to be determined is already specified by entering limit values into the user application.

[0014] A further aspect of the invention relates to a computer program configured to carry out the above-described method using a computer. It should be understood that the computer is connected to the camera-based sensor via an interface, so that the images captured by the camera-based sensor can be processed in the computer.

[0015] A further aspect of the invention relates to a control device for a luminaire with a camera-based sensor, wherein the control device is configured to automatically carry out the previously described method. The control device can be integrated together with the camera-based sensor at the location of the luminaire or in the luminaire itself. However, it is also conceivable to accommodate the camera-based sensor together with the control device in one location and to connect the luminaire or, if applicable, several luminaires to the control device. For example, a camera-based sensor can be provided in an office space, which controls several luminaires in the office space. It is also possible for a camera-based sensor to control several traffic lights. However, the one or more luminaires to be controlled is / are preferably arranged within sight of the camera-based sensor.

[0016] Further advantages and features of the present invention will become apparent from the following description of preferred embodiments, which is given in conjunction with the figures. The figures depict the following. Fig. 1 shows a flowchart of a method according to the present invention. Figs. 2A-2C show a top view of an office space, the same top view with movement profiles, and the same top view with a permitted movement area.

[0017] Referring to Fig. 1 A method according to one embodiment of the present invention is shown. The method is carried out fully automatically in a control device of a camera-based sensor of a luminaire.

[0018] In step S10, the method includes recording the positions (in particular, X, Y values in an area) and an activity level of each object within a detection range of the camera-based sensor. The object can be, for example, a person or any other moving object. An activity level for each object is determined based on the frequency of the respective object's position changes.

[0019] In step S20, light values detected by the camera-based sensor in the detection area are recorded.

[0020] The recorded positions, activity levels, and light values are stored as data records in a memory. Storing these data records for the positions (X, Y) is also referred to below as heatmap creation.

[0021] In step S30, which is purely optional, the heatmap image is selected depending on a customer application. For this purpose, a customer can use a computer-assisted user application to enter threshold values that are taken into account when recording the data records. Data records at positions outside the defined threshold values are discarded during recording.

[0022] In the next step S40, the heatmap of the object position is recorded in combination with the measured light values.

[0023] The preceding steps are performed multiple times in succession until an analysis is performed in step S50, which is performed every minute or after every 1,000 data sets. The analysis may include a plausibility check. In particular, position jumps above a threshold or unusual trajectories can be detected, and the corresponding data sets are discarded. Furthermore, the object positions are analyzed with regard to the duration of the position and the change in position in order to determine an activity level for the object. Finally, the change in light value is analyzed, and the object positions are compared. Based on these analyses, the heatmap data is adjusted.

[0024] The preceding steps are performed several times in succession until an analysis is performed in step S60 after 50,000 data records or after 24 hours. During this analysis, implausible data records are deleted. Furthermore, individual data points are analyzed using parameter sets based on the previously defined thresholds in the user application. Data records outside the defined thresholds are also discarded. The heatmap data is adjusted according to the analysis.

[0025] The preceding steps are performed several times in succession until a final assessment of the plausibility of the recorded heatmap is performed in step S70 after 100,000 data sets or after 48 hours. During the assessment, edges are formed from the active object movements to calculate regions. This adjusts the heatmap data and generates active and inactive regions. The active regions are image areas in which object movements with an activity level above a specified threshold were recorded, while the remaining regions are stored as inactive movement regions. Optionally, the regions can be formed by regular or irregular polygons with smooth edges.

[0026] In step S80, the plausibility of the heatmap is checked. The plausibility assessment is carried out based on recorded movement data of the objects. Object movements with position jumps above a defined limit or object movements with a trajectory that deviates from a defined norm, e.g., with regard to straightness or continuity of movement, are marked as implausible movement data. If a limit for implausible movement data is exceeded, the result of the heatmap is determined to be implausible, and the process is repeated from step S10. If the plausibility assessment is positive, the active and inactive movement areas are saved in step S90. For the subsequent control operation of the luminaire, only the active movement areas are monitored by the camera-based sensor system in order to use this monitoring to control the luminaire.

[0027] The Figuren 2A bis 2C show an example of how an area to be illuminated can be set up when setting up the camera-based sensor of the luminaire. Fig. 2A shows an office room in plan view, in which two desks 2, two desk chairs 1, two shelves 3 and a conveyor belt 4 are arranged in a detection area 5 (corresponds to the maximum possible detection range of the camera-based sensor). Fig. 2B shows the same supervision as Fig. 2A , with a recorded movement profile 6 additionally shown. It can be seen that the movement profile 6 extends predominantly to the areas provided between the furnishings of the office space. Furthermore, it can be seen that movements above the desk chairs 1, and on partial areas of one of the desks 2 and the conveyor belt 4 were also recorded. However, the movement profile 6 only represents the recorded movement over a limited period of time. For the analysis, a large number of such movement profiles are recorded, and the activity levels of the movement profiles are determined for each of them. The active movement areas are determined from the areas with movement profiles and an increased activity level. Fig. 2Cthe active movement area 7 is shown. The movement area includes all areas of the office space outside of the restricted areas 8. In this example, the restricted areas are the areas of the shelves 3 and the desks 2 as well as the area of the conveyor belt 4 and the area of the conveyor belt 4 facing away from the shelves 3 and the desks 2. In these areas, no movements or only movements with a low activity level were detected. Accordingly, they are saved as inactive movement areas, while the areas with an increased activity level are saved as active movement areas 7. When the office lighting is operated, which can be by one or more luminaires, only the active movement areas 7 are monitored by means of the camera-based sensor system.The remaining areas are not monitored, so that activities in these areas do not result in any change in the lighting level of the luminaire(s) due to the control via the camera-based sensor.

Claims

1. A method for setting up a camera-based sensor for a luminaire, comprising: commissioning the camera-based sensor at the location of the luminaire, recording positions and movements of objects by means of the camera-based sensor in a detection range of the camera-based sensor, analyzing the recorded object positions and object movements with respect to a frequency of position changes of the objects, and determining an activity level of the objects from the frequency of position changes, recording the position and activity level of each object, determining active and inactive movement areas from the object positions and the activity levels, wherein areas in which object positions of objects with an activity level above a predetermined limit were recorded are defined as active movement areas,and the remaining areas of the detection range of the camera-based sensor are defined as non-active motion areas, and limiting the motion detection by the camera-based sensor system when operating the luminaire to the active motion areas.

2. The method according to claim 1, wherein the step of analyzing the recorded object positions and movements with respect to the frequency of position changes of the objects is repeated at a time interval of at least one minute or after recording at least 1,000 data sets concerning the position and movement of objects, and the step of recording the position and activity level of each object is adjusted according to the repeated analysis.

3. The method according to claim 2, wherein during the repeated analysis of the recorded object positions and movements, an automatic plausibility check is additionally carried out, which determines position jumps of the object based on the recorded movement data of the object, and the method further comprises: deleting recorded data records that contain implausible movement data of an object, wherein implausible movement data are movement data in which position jumps above a predetermined limit value have been detected, and updating the recordings of the positions and the activity level of each object only taking into account the non-deleted data records.

4. The method according to claim 3, wherein the deletion of the data records containing implausible movement data of an object occurs after at least 24 hours or after recording at least 50,000 data records concerning the position and movement of objects by means of the camera-based sensor.

5. Method according to one of the preceding claims, wherein the step of determining active and non-active movement areas from the object positions and the activity levels is carried out after at least 48 hours or after at least 100,000 recorded data sets concerning the position and movement of objects by means of the camera-based sensor.

6. The method according to any one of the preceding claims, wherein the step of recording positions and movements of objects using the camera-based sensor further comprises: recording light values in the detection area using the camera-based sensor; and after at least one minute or after at least 1,000 recorded data sets of the light values, analyzing changes in the light values and comparing them with recorded object positions.

7. The method according to any one of the preceding claims, wherein the method further comprises: providing a user application configured to: input threshold values used in the analysis of the recorded object positions and movements in order to discard data records that exceed the defined threshold values.

8. A computer program comprising instructions which, when executed on a computer in conjunction with a camera-based sensor, carry out the method according to any one of the preceding claims.

9. Control device of a luminaire with a camera-based sensor, wherein the control device is configured to carry out the method according to one of claims 1 to 7.

Images