Area classification with target tracking

Abstract

A device to identify and classify an area in a space are provided. The device includes a light source to emit light to illuminate a space. In addition, the device includes a light source controller to control the light source. The device also includes a sensor to measure light data in the space. Furthermore, the device includes an image processing engine to track a plurality of targets across a field of view. The device includes a memory storage unit to store the local tracking data for the plurality of targets, and an aggregator to analyze the local tracking data and to generate a classification of an area of the space. A system to illuminate a space with a plurality of the devices disposed in a space is also provided. In addition, a method of identifying and classifying an area in a space are provided.

Description

AREA CLASSIFICATION WITH TARGET TRACKINGBACKGROUND

[0001] Buildings typically have spaces which may be used for varying purposes. For example, some rooms may be used as a general meeting room where several individuals may congregate to facilitate communication, such as for a meeting. As another example, some rooms may be used as a private office which may be assigned to one individual at a time, where the individual may have privacy to improve concentration. Other types of spaces may include open spaces with high ceilings and partial barriers such as furniture or shelving. Accordingly, rooms may have a variety of sizes and shapes and are typically separated by a boundary or obstacles.SUMMARY

[0002] This invention is a mechanism for a lighting control system to automatically detect, configure, describe one or more area or room types in a completely automatic manner. It reduces pre-installation planning, onsite programming, or any other type of setup to manually assign each luminaire to the appropriate area or room type.

[0003] The invention fuses endogenous lighting system information with exogenous information from the environment and the interactions of objects therein. Endogenous information may include the direct, indirect, or reflected electro-optical signal emitted a device and received by another device. Exogenous information may include the spatial and temporal motion patterns detected by each device. The fusing of endogenous and exogenous information can allow each device to: A) determine membership in a set of co-located devices, B) determine the type of area or room where they have been installed, C) determine membership in a subset of devices that should synchronize operating states, D) synchronize operation with the correct subset of devices, and E)describe the utilization of the subset of devices.

[0004] In accordance with an aspect of the invention, there is provided a device. The device includes a light source to emit light to illuminate a space. In addition, the device includes a light source controller to control the light source. The light source controller is to change an intensity of the light emitted by the light source. Furthermore, the device includes a sensor to measure light data in the space. The light data includes target information. The device also includes an image processing engine to track a plurality of targets across a field of view. The image processing engine generates local tracking data for each target of the plurality of targets. The device further includes a memory storage unit to store the local tracking data for the plurality of targets. The device also includes an aggregator to analyze the local tracking data and to generate a classification of an area of the space. The light source controller uses the classification to control the light source.

[0005] In accordance with an aspect of the invention, there is provided a system. The system includes a plurality of lighting devices disposed in a space. Each lighting device of the plurality of lighting devices includes a light source to emit light. Each lighting device also includes a light source controller to control the light source. The light source controller of each lighting device is to change an intensity of the light emitted by the light source. Furthermore, each lighting device includes a sensor to measure light data in the space. The light data includes target information. In addition, each lighting device includes an image processing engine to track a plurality of targets across a field of view. The image processing engine of each lighting device generates local tracking data for each target of the plurality of targets. Additionally, each lighting device includes a communications interface to communicate with other lighting devices of the plurality of lighting devices. The communications interface of each lighting device receives external tracking data from the other lighting devices. The communications interface of each lighting device also shares the local tracking data with the other lighting devices. Each lighting device also includes an aggregator to analyze the local tracking data and the external tracking data to group a subset of the plurality of lighting devices. The subset is communicated to other devices. The system also includes a system controller to control the plurality of lighting devices, wherein the subset of theplurality of lighting devices is to be controlled together.

[0006] In accordance with an aspect of the invention, there is provided a method of classifying an area in a space. The method involves measuring light data in a space with a sensor. The light data includes target information. In addition, the method involves tracking a plurality of targets across a field of view of the sensor with an image processing engine. Furthermore, the method involves generating local tracking data for each target of the plurality of targets. The method also involves analyzing the local tracking data to generate a classification of an area of the space. The method additionally involves controlling a light source to change an intensity of light emitted by the light source based on the classification.BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Reference will now be made, by way of example only, to the accompanying drawings in which:

[0008] Figure 1 is a schematic representation of the components of an example of a device to identify and classify an area in a space;

[0009] Figure 2 is a schematic representation of the components of another example of a device to identify and classify an area in a space;

[0010] Figure 3 is a schematic representation of the components of another example of a device to identify and classify an area in a space;

[0011] Figure 4 is a schematic representation of an example of a system that to illuminate a space;

[0012] Figure 5 is a schematic representation of the example system shown in figure 4 with a plurality of targets moving through the space;

[0013] Figure 6 shows the field of view of a device tracking a target moving therethrough and the data generated by the image processing engine;

[0014] Figure 7 shows the field of view of a device tracking multiple target movements and the data generated by the image processing engine;

[0015] Figure 8 is a schematic representation of the example system shown in figure 4 with a target moving through the space into an aisle;

[0016] Figure 9A shows the field of view of a device tracking the movement of target in the example show in figure 8 and the tracking data collected by the device;

[0017] Figure 9B shows the field of view of another device tracking the movement of target in the example show in figure 8 and the tracking data collected by the device;

[0018] Figure 9C shows the field of view of another device tracking the movement of target in the example show in figure 8 and the tracking data collected by the device;

[0019] Figure 10 is a schematic representation of the example system shown in figure 8 monitoring multiple paths in the space;

[0020] Figure 11 A is a schematic representation of another example system monitoring multiple paths in the space;

[0021] Figure 11 B is a schematic representation of the example system shown in figure 11A grouping devices;

[0022] Figure 12 is a flowchart of an example of a method of classifying an area in a space; and

[0023] Figure 13 is a schematic representation of the components of another example of a device to identify and classify an area in a space.DETAILED DESCRIPTION

[0024] In the following detailed description, certain examples of the system will be described, however those skilled in the art will appreciate that variations exist and can be implemented in a manner that is consistent with the novel aspects of this disclosure.

[0025] Room types that are manually specified by user intervention during the commissioning phase of the lighting system do not adapt to changes in how the space is used without further user intervention. For example, a storage room might be repurposed as a private office, or an enclosed office area might be modified into an open plan after the manual specification is made. In another example, a retail store may call for the lights above a given aisle are controlled together and independently of the lights in other aisles or other parts of the store. It is to be understood that such a simple motion detection scheme ignores the nuanced patterns of information that is available from the spatial and temporal tracking of objects throughout the area. For example, a simple motion detection scheme cannot distinguish between concurrent motion activity in the same or in different aisles of a retail store.

[0026] In some examples, a luminaire-mount sensor equipped with an active imaging subsystem and multiple target tracker that works with other luminaire-mount sensors to collectively determine the type of area or room in which they are co-located. A system to classify the aisle or room includes a near infrared emitter, an electro-optical focal plane array, a mesh radio, and a processing subsystem. The processing subsystem is capable of distributed resource management for active imaging among an array of such systems. In addition, the processing subsystem is capable of spatial feature detection (e.g. variation in digital signals across the focal plane array). Furthermore, the processing subsystem is capable of temporal feature detection (e.g. changing motion, ambient light level, various physical phenomena over time). The processing subsystem is also capable of both spatial and temporal multiple target tracking (i.e. estimation of the state of objects through space and time) and distributed inference among an array of such systems (including the room type). The processing subsystem may use the information collected to aggregate statics that describe how the area is used (e.g. flow, people counting, ingress / egress points, etc.).

[0027] It is then appreciated by a person of skill with the benefit of this description that the system may determine the spatial and temporal signals which are endogenous and exogenous to the lighting system; share and compare these signals among an array of sensors; classify the room type / area type based on the shared information; and describe the utilization of the area.

[0028] A system is provided to receive and collect data from an area, such as a room. The manner by which the data is received is not limited and may include receiving information directly from sensors. In other examples the data may be received from other devices, such as over a network. The data received by the system may include ambient light level, infrared active imaging schedule, focal plane array data, time of day, processor runtime, luminaire active power, and radio telemetry.

[0029] The system samples various signals and pre-filters and amplifies to increase the signal to noise levels. Features are then extracted from the focal plane array based on spatial and temporal image processing techniques. Feature-to-track assignments may be made to support multiple target tracking and multiple-state tracking filters continuously estimate a wide spectrum of target attributes. The endogenous and exogenous features are compressed into their principal representations and shared over a mesh network. Neighboring nodes fuse this information into a normalized feature set for classification.

[0030] A supervised machine learning model may be trained and deployed into the lighting system nodes. The supervised machine learning model periodically runs an inference on the feature set. Training is carried out using a diverse set of actual lighting installations. Data augmentation is used to reduce bias, increase generalization, and to allow for domain adoption across various room types. Several locations were used to validate the results of the algorithms.

[0031] The system may be used to classify room types to a high level or precision and recall. In the present examples, the classification process includes many unique and innovative aspects. For example, by observing the exogenous signals in the environment (including interactions between objects), the lighting system can learn how to let the environment declare its type and how it is being utilized. By correlating these signals with data inherent to the lighting system, the system can fine-tune responses in a manner that is human-centric.

[0032] Referring to figure 1 , a schematic representation of a device to identify and classify an area in a space is generally shown at 50. The device 50 may be a lighting fixture and may include additional components, such as various additional interfaces and / or input / output devices such as indicators to interact with a user of the device 50.The interactions may include viewing the operational status, updating parameters, or resetting the device 50. In the present example, the device 50 is to measure and collect data to map a space and to classify certain areas and the use thereof. For example, objects or other obstacles in a space may affect the manner by which the space is used. In particular, if an obstacle, such as a partition or piece of furniture is placed in a room, users of the room generally move around the obstacle. In the present example, the device 50 includes a light source 55, a light source controller 60, a sensor 65, an image processing engine 70, a memory storage unit 75, and an aggregator 80.

[0033] The light source 55 is to emit light for illuminating a space. The light source 55 is not particularly limited and may be any device capable of generating light for the purposes of space lighting. For example, the light source 55 may be an incandescent light bulb, a fluorescent light bulb, or a light emitting diode.

[0034] In the present example, the light source controller 60 is to control the light source 55. In particular, the light source controller 60 may provide power to the light source 55 to control the intensity of the light emitted by the light source 55. The light source controller 60 may also turn off the light source 55 by cutting power when lighting in the area proximate to the device 50 can be reduced, such as due to low activity in the area. In some examples, the light source controller 60 may also control the dimming of the light source 55 as well as dimming rates and other operating conditions. The manner by which the light source controller 60 controls the light source 55 is not particularly limited and may involve implementing appropriate operating parameters based on a classification, as determined by the aggregator 80, of the area in which the device 50 is located. In further examples, the light source controller 60 may receive data from an external source or a peripheral sensor to control the light source 55 based on various conditions in the space, such as the ambient light level, motion, or time. It is also to be understood by a person of skill with the benefit of this example that the light source controller 60 may control the intensity based on commands received from a user interface.

[0035] The sensor 65 is to measure light data in the space. In particular, the sensor 65 may be used to collect data that includes information about a target in the field of view of the sensor 65. The target is not particularly limited and may be a person orother moving object in the field of view. The sensor 65 is not particularly limited and may be any sensor capable of measuring light data, such as a camera. In the present example, the sensor 65 is capable of capturing two-dimensional image in the visible spectrum and may also detect light in the infrared spectrum with a wavelength of about 850 nm or longer. In some examples, a lens may be used to provide a wider coverage area to increase a field of view to detect motion patterns of targets.

[0036] In other examples, the sensor 65 may be a low resolution sensor. In particular, the sensor 65 may have a resolution sufficiently low such that the light data captured is cannot be used to distinguish or identify people. However, the sensor 65 may have sufficient resolution to identify targets in the field of view. In addition, movement patterns of targets within the field of view may tracked as the target moves within the field of view of the sensor 65. The number of pixels in a low resolution sensor is not particularly limited and may be sufficiently low to not store identifying features of the targets. For example, each sensor 65 may have as few as four pixels across to cover a field of view of about 20 m. Other examples may have eight pixels, sixteen pixels across to cover a field of view of about 20 m. In other examples, the sensor may have more pixels to improve detection and tracking of objects, but not to provide the capability to distinguish identifying features of the target.

[0037] The image processing engine 70 is process the light data collected by the sensor 65 to identify a plurality of targets and to track the targets moving across the field of view of the sensor 65. The manner by which the image processing engine 70 processes the light data is not particularly limited. Based on the target or targets identified in the light data, the image processing engine 70 generates tracking data for each target that describes the motion of the target through the field of view. In some examples, the image processing engine 70 may use image recognition techniques to identify a target moving through the field of view of the sensor 65. In particular, the image processing engine 70 may analyze the light data for changes in spatial signals that may be indicative of a target moving through the field of view. The changes in the spatial signal may then be used to identify the target moving through the field of view. It is to be appreciated that in some examples, multiple targets may be in the field of view and the sensor 65 to be identified. In some examples, the image processing engine 70may use artificial intelligence or machine learning techniques to identify each separate target when there are multiple targets in the field of view. It is to be appreciated by a person of skill with the benefit of this description that the tracking data generated locally by the image processing engine 70 may include data describing the path of the target through the field of view. In examples with multiple targets being tracked, the image processing engine 70 may generate tracking data with multiple paths.

[0038] The memory storage unit 75 is to store the tracking data generated by the image processing engine 70. In the present example, the memory storage unit 75 may continuously collect tracking data from the image processing engine 70 to be provide to the aggregator 80. The tracking data may be collected over a period of time, such as a few hours or a few days. Accordingly, the stored tracking data may include the paths of many targets passing through the field of view of the sensor 65.

[0039] The memory storage unit 75 may be also used to store addition data to be used by the device 50. For example, the memory storage unit 75 may store ambient light data, or data received from external devices as discussed in greater detail below. Furthermore, the memory storage unit 75 may be used to store mapping data and information other information from external devices which may be adjacent or proximate.

[0040] In the present example, the memory storage unit 75 may include a non- transitory machine-readable storage medium that may be any electronic, magnetic, optical, or other physical storage device. In other examples, the memory storage unit 75 may be an external unit such as an external hard drive, or a cloud service providing content. The memory storage unit 75 may also be used to store instructions for general operation of the device 50. In particular, the memory storage unit 75 may store an operating system that is executable by a processor to provide general functionality to the device 50, for example, functionality to support various applications. The memory storage unit 75 may additionally store instructions to operate the image processing engine 70 or the aggregator 80. Furthermore, the memory storage unit 75 may also store control instructions to operate other components and peripheral devices, such additional sensors, cameras, user interfaces, and light sources.

[0041] The aggregator 80 is in communication with the memory storage unit 75 fromwhich the aggregator 80 receives the tracking data to analyze. The aggregator 80 is to generate a classification of an area proximate to the device 50, which can be used by the light source controller 60 to control the operation of the light source 55. The manner by which the aggregator 80 generates the classification may involve analyzing paths from the tracking data. The paths may indicate the presence of objects in the space, which in turn may be used to generate the classification. In the present example, the space may be a business with a high ceiling and multiple shelves that are about two meters tall. In particular, the space may be a store with shelving on which products are placed. In this example, the targets may be customers walking through the store.Since the customers generally walk down each aisle instead of through the shelving, the paths stored in the memory storage unit 75 may be used to determine the locations of the shelves and to classify areas of the space as being an aisle. Based on the classification of the space, the aggregator 80 may be able to identify locations in the space where an object is placed. It is to be appreciated by a person of skill with the benefit of this description that the identification of areas in the field of view may be used to generate a utilization map of the space.

[0042] Referring to figure 2, another schematic representation of another device to identify and classify an area in a space is generally shown at 50a. Like components of the device 50a bear like reference to their counterparts in the device 50, except followed by the suffix “a”. In the present example, the device 50a is to communicate with other external devices to receive additional tracking data. In the present example, the device 50a includes light source 55a, a light source controller 60a, a sensor 65a, an image processing engine 70a, a memory storage unit 75a, an aggregator 80a, and a communications interface 85a.

[0043] In the present example, the communications interface 85a is to communicate with an external device (not shown). In the present example, the communications interface 85a may communicate with external devices over a network, which may be a public network shared with a large number of connected devices, such as a WiFi network or cellular network. In other examples, the communications interface 85a may be to communicate over a private network via a Bluetooth connection, radio signals or infrared signals. In particular, the communications interface 85a may communicate withan external device to receive tracking data from the external device. In the present example, the external device may be similar or identical to the device 50a, wherein the external device includes an analogous aggregator to generate tracking data based on data collected by the external device. It is to be appreciated by a person of skill with the benefit of this description that multiple devices may be place in a space, such as on a ceiling to operate as light fixtures to collect light data. Accordingly, each of the devices in the space may generate tracking data from the collected light data, which is to be received by the communications interface 85a. The received tracking data may then be stored on the memory storage unit 75a.

[0044] The aggregator 80a is in communication with the memory storage unit 75a from which the aggregator 80a receives the tracking data collected locally on the device 50a and the tracking data received from external devices via the communications interface 85a to analyze. The aggregator 80a combines both sources of tracking data to generate a classification of an area proximate to the device 50a, which can be used by the light source controller 60a to control the operation of the light source 55a. The manner by which the aggregator 80a combines the tracking data is not particularly limited. In the present example, the local tracking data and external tracking data may have portions of overlapping paths which can be match across multiple fields of view of the device 50a and the external devices. As the field of view of the sensor 65a is generally offset from the field of view of corresponding sensors of the external devices, the aggregator 80a factors in the offset to combine the various fields of view to match paths from different devices. The combination will provide longer paths with which the analysis can be carried out that may extend beyond the field of view of the sensor 65a. Combining and matching the external tracking data with the local tracking data can involve using known relative locations of each external device from which the communications interface 85a receives data as well as the timestamps for each path in the tracking data. The relative locations of each external device to the device 50a may be determined using various localization methods. For example, additional components of the device 50a, such as infrared light sources and sensors, ultrasonic emitters and receivers, or other proximity sensing systems may be used triangulate the relative positions of each external device. In other examples, the relative positions may bepreprogrammed into the device 50a during the installation process.

[0045] In some further examples, the communications interface 85a may also be configured to transmit the tracking data generated by the image processing engine 70a to external devices. In this example, the manner by which the communications interface 85a transmits and receives the data is not limited and may include receiving an electrical signal via a wired connection with other external devices or via a central server.

[0046] Referring to figure 3, another schematic representation of a device to identify and classify an area in a space is generally shown at 50b. Like components of the device 50b bear like reference to their counterparts in the device 50a, except followed by the suffix “b”. In the present example, the device 50b is to communicate with other external devices to receive additional tracking data and to assign groups of devices to operate as together. In the present example, the device 50b includes light source 55b, a light source controller 60b, a sensor 65b, an image processing engine 70b, a memory storage unit 75b, an aggregator 80b, a communications interface 85b, and a grouping engine 90b.

[0047] The grouping engine 90b is not particularly limited and may be operated by a common processor, a separate processor or even a separate machine, such as a central controller, in other examples. The grouping engine 90b is to associate the device 50b with a plurality of external devices autonomously. In the present example, the devices that cover and area with a common classification may be associated as a group to be controlled together. For example, a classification of an area may be an aisle in the store between shelves. In this example, the area of the space may be the aisle. Accordingly, the devices within the aisle may be associated together such that devices in the aisle may be activated together, such as turning on the lighting devices within the aisle may be controlled together.

[0048] It is to be appreciated by a person of skill with the benefit of this description that the grouping engine 90b may continuously operate to associate the device 50a with different external devices. In some examples, the grouping engine 90b may review the area classifications periodically, such as each day or each week. In other examples, the grouping engine 90b may be prompted with a command received via thecommunications interface 85b to re-evaluate associations with external devices. This functionality allows a space to be reconfigured without reprogramming groupings for different areas within the space. Furthermore, in some examples, the grouping engine 90b may share grouping data with the associations of the device 50a with other devices to compare and confirm the grouping data or to extend the grouping beyond the range of the communications interface 85b.

[0049] Referring to figures 4 and 5, a schematic representation of a system 100 that is configured to illuminate a space 200 is generally shown. In the present example, the system 100 include a plurality of devices 50a-1 , 50a-2, 50a-3, 50a-4, 50a-5, 50a-6, 50a-7, 50a-8, 50a-9, 50a-10, 50a-11 , 50a-12, 50a-13, 50a-14, 50a-15, and 50a-16 (generically, these devices are referred to herein as “device 50a” and collectively they are referred to as “devices 50a”) deployed in the space 200. The system 100 also includes a system controller 110.

[0050] In the present example, the space 200 can be assumed to be a store, such as a grocery store. The space 200 includes aisles 205-1 , 205-2, 205-3, and 205-4 (generically, these aisles are referred to herein as “aisle 205” and collectively they are referred to as “aisles 205”). The aisles 205 are separated by shelves 210-1 , 210-2, 210-3, 210-4, and 210-5 (generically, these shelves are referred to herein as “shelf 210” and collectively they are referred to as “shelves 210”). In this example, the shelves 210 may be about two meters tall such that a person (i.e. the target 220) can comfortably reach items on the shelves 210. The space 200 may have a higher ceiling of five meters or more, which is typical of many retail spaces. To provide illumination, the devices 50a are generally installed close to the ceiling and well above the tops of the shelves 210. Therefore, there is no barrier between adjacent devices 50a that can be used to infer the use of the space 200. In the present example, the space 200 also includes an office 215 over which device 50a-4 is disposed. The office 215 is not particularly limited and may include a sealed space with a wall extending to the ceiling of the space 200 or may be partitioned with a wall or other barrier that extends a few meters above the floor such that the device 50a-4 is also not isolated from adjacent devices 50a-3, 50a-7, and 50a-8.

[0051] In the present example, each of the devices 50a are substantially identicalunits and operate together to illuminate the space 200 collectively or individually. The devices 50a are lighting devices that were described in greater detail above. In the present example, the devices 50a are generally spaced uniformly throughout the space 200 and the field of view 51 a of the sensor 65a on each device 50a overlaps with at least the field of view 51a of the sensor 65a on an adjacent device 50a such that the entire space 200 is covered by the plurality of devices 50a such that targets 220 can be tracked as they move through the space 200.

[0052] The system controller 110 is to control the devices 50a. The manner by which the system controller 110 controls the devices 50a is not particularly limited. For example, the system controller 110 may include a user interface, such as a touchscreen device, to receive user input to operate each of the devices 50a. The system controller 110 may operate each device 50a individually to adjust operating parameters, such as the intensity of light to be emitted by the light source 55a. In addition, for devices 50a that have been grouped into subsets, the system controller 110 may be used to control the group of devices 50a. In the present example of the system 100, it is to be appreciated by a person of skill with the benefit of this description that the grouping is not particularly limited. In some examples, the grouping may be carried out by the system controller 110 based on data receive from the aggregator 80a of each device 50a. For example, devices 50a covering a specific area of the space may be grouped to providing lighting. Continuing with the present example, device 50a-5 and device 50a-6 may be grouped together in a subset to illuminate aisle 205-2 to allow a person (i.e. target 220) to walk down an illuminated aisle and see items on shelf 210-2 and shelf 210-3 without powering on additional devices 50a in the space 200 that are further from the aisle 205-2. Other subsets may also be formed to illuminate the other aisles 205 of the space. It is to be appreciated that this manner of illuminating the aisle 205-2 is advantageous than to simply activating a device 50a above the target 220, such as with a simple motion sensor, as the target 220 may not be able to see items further down the aisle before walking closer.

[0053] In other examples, each device 50a may include a grouping engine to determine groupings locally on each device 50a. In this example, the groupings may be transmitted to the system controller 110. Furthermore, where each device 50a includesa grouping engine, the device 50a may further include a voting engine to compare groupings from multiple devices 50a to determine a consensus on the subset of devices 50a to for each group.

[0054] In further examples, the aggregator 80a of each device 50a may generate a utilization map of an area of the space 200 proximate to the device 50a. The utilization maps from each device 50a may then be transmitted to the system controller 110 and analyzed to generate an aggregated map of utilization across the space 200. The aggregated map may then be transmitted to a user, displayed on a user interface device, or stored for subsequent use.

[0055] It is to be appreciated by a person of skill with the benefit of this description that alternative methods from path tracking to determine utilization may be implemented. For example, one or more device 50a may be equipped with an additional counting device (not shown) to count targets passing near the device. In this example, a determination of the number of hits by the counter may indicate the utilization of the area proximate to the device 50a.

[0056] Figure 6 illustrates generation of tracking data by the image processing engine 70a to identify the path 250 of a target 220 through the field of view 51a of a device 50a. In the present example, the image processing engine 70a detects the entry of the target 220 into the field of view 51a and records a TRACK INIT datapoint in the tracking data. The image processing engine 70a then samples the light data received from the sensor 65a periodically, such as every second. As the target 220 moves through the field of view 51a, the image processing engine 70a may detect a new location of the target to record TRACK LOCK datapoints. At some sample points, the image processing engine 70a may fail to detect the target 220 in the field of view. In this case, the image processing engine 70a records TRACK COAST datapoints which extrapolate the location of the target 220 based on the last known location and the last known speed of the target 220. If the image processing engine 70a reacquires the lock on the target 220, the image processing engine 70a continues to record TRACK LOCK datapoints. As the target hits the boundary of the field of view 51a, the image processing engine 70a records a TRACK TERMINATION datapoint as the device 50a will no longer be able to track the target 220. After the target 220 leaves the field ofview 51a, the image processing engine 70a calculates the path 250 based on the tracking datapoints. Figure 7 illustrates the image processing engine 70a processing multiple targets passing through the field of view to generate a plurality of paths.

[0057] Referring to figure 8, an example analyzing the movements of a target 220 walking through the space 200 is shown. In this example, the target 220 moves through an open space and into the aisle 205-2 along the path 250. Figure 9A illustrates the field of view 51a of the device 50a-7 which detects the path 250. Figure 9B illustrates the field of view 51 a of the device 50a-6 which detects the path 250. Figure 9B illustrates the field of view 51a of the device 50a-5 which detects the path 250. It is to be appreciated by a person of skill with the benefit of this description that other devices may also detect a portion of the path.

[0058] During operation, figure 10 shows the system 100 monitoring multiple paths through the space 200. It is to be appreciated by the paths that an obstacle can be inferred between aisle 205-2 and aisle 205-3. This determination may be used to group devices 50a-5 with device 50a-6 as a subset. Similarly, device 50a-9 and 50a-10 may be groups as another subset.

[0059] Figures 11 A and 11 B show the space 200 reconfigured to have aisles 305-1 , 305-2, and 305-3. In this example, shelf 310-1 and shelf 310-2 are used to form the aisles 305 (generically, these aisles are referred to herein as “aisle 205” and collectively they are referred to as “aisles 205”). As discussed above, the system 100 may automatically reconfigure the subsets of devices in this configuration to include devices 50a-1 , 50a-5, and 50a-9 as a subset after sufficient targets move through the aisles 305 in this configuration. Similarly, devices 50a-2, 50a-6 and 50a-10 may be grouped as another subset and devices 50a-3, 50a-7 and 50a-11 may be grouped as another subset.

[0060] In the present example, the determination of the grouping is not particularly limited. In one example of determining the grouping of the devices 50a, each device 50a may analyze the paths 250 of targets passing within proximity of the device. As a specific example, device 50a-6 may track multiple paths 250 passing within proximity. In this example, it is assumed that the device 50a-6 has identified the closes devices to be device 50a-5, device 50a-10, device 50a-7, and device 50a-2. The identification ofthe devices 50a is not limited and may be preprogrammed or determine based on various localization methods. Therefore, the device 50a-6 may assume that one of device 50a-5, device 50a-10, device 50a-7, and device 50a-2 is to be included in the same group.

[0061] Accordingly, the device 50a-6 may analyze the tracking data from the image processing engine 70a to make the determination. In one example of the analysis, the device 50a-6 may calculate an aggregate path 255 which may be an average of the paths moving through the field of view 51a of the device 50a-6. The average may be calculated as the midpoint of two paths 250 at the same after the target enters the field of view. Upon determining the aggregate path 255, the device 50a-6 may calculate the angle 0 between the aggregate path 255 and the axis 405 of the devices 50a-2 and 50a-10, and the angle ( between the aggregate path 255 and the axis 410 of the devices 50a-5 and 50a-7. The device 50a-6 may group itself with the devices with the smaller angle or with an angle below a certain threshold value. In the present example, the angle 0 is smaller than the angle ( . Therefore, the device 50a-6 may group itself with device 50a-2 and 50a-10.

[0062] Referring to figure 12, a flowchart of an example method of classifying an area in a space is generally shown at 500. In order to assist in the explanation of method 500, it will be assumed that method 500 may be performed with the device 50. Indeed, the method 500 may be one way in which the device 50 may be configured. Furthermore, the following discussion of method 500 may lead to a further understanding of the device 50 and its components. In addition, it is to be emphasized, that method 500 may not be performed in the exact sequence as shown, and various blocks may be performed in parallel rather than in sequence, or in a different sequence altogether.

[0063] Beginning at block 510, light data in a space is to be measured with a sensor 65. The manner by which the light data is measured is not particularly limited. In the present example, the sensor may be a camera to detect light within a field of view. The camera may detect a spatial feature within the field of view as well as a timestamp of when the light data was measured. It is to be appreciated that information about a target, such as its location and movement may be in the light data.

[0064] Block 520 comprises tracking a plurality of targets in the light data measured at block 510. In the present example, the light data from the field of view of the sensor 65 of the device 50 is analyzed with the image processing engine 70 to ultimately generate tracking data at block 530.

[0065] Next, block 540 comprises analyzing the tracking data generated at block 530 to assign a classification of the area proximate to the device 50. The classification may then be used by the light source controller 60 to change the intensity of the light emitted by the light source 55. In particular, the light source 55 may be controlled as part of a group of devices such that the area of the space may be illuminated while leaving other areas of the space that are not utilized to be not illuminated to save energy.

[0066] In some examples, block 540 may also include receiving tracking data from external sources, such as similar neighboring devices. The received tracking data may be combined with the tracking data generated locally by the image processing engine 70 to provide more information to make the classification of the area proximate to the device 50. Similarly, the device 50 may operate as an external device to an adjacent device such that the device 50 may transmit the tracking data generated by the image processing engine 70 to an external device.

[0067] Referring to figure 13, another schematic representation of a device to identify and classify an area in a space is generally shown at 50c. Like components of the device 50c bear like reference to their counterparts in the device 50b, except followed by the suffix “c”. In the present example, the device 50c is to communicate with other external devices to receive additional tracking data and to assign groups of devices to operate as together. In the present example, the device 50c includes light source 55c, and infrared source 57c, a sensor 65c, an image processing engine 70c, a memory storage unit 75c, an aggregator 80c, a communications interface 85c, and a processor 95c.

[0068] The infrared source 57c is to emit infrared light. The light may be monochromatic or emit a band of light with a peak wavelength in the infrared spectrum. For example, the infrared source 57c may emit light having a peak wavelength greater than about 780 nm to be beyond the typical visual range of a human eye. In some examples, the peak wavelength may be about 850 nm. The infrared source 57c is notparticularly limited and may be any device capable of generating light that may be reflected off a surface, such as a room boundary, and detected by an infrared sensor on a proximate device. For example, the infrared source 57c may be an incandescent light bulb, a fluorescent light bulb, a laser, or a light emitting diode. The area onto which the infrared source 57c projects is not particularly limited. In the present example, the infrared source 57c may project a uniform intensity t be detected by a corresponding sensor. In other examples, the infrared source 57c may direct a wider or narrower light, or the illumination may not be uniform across substantially the field of view. In the present example, the infrared source 57c is used to determine if there is a boundary, such as a wall between the device 50c and a proximate device. It is to be appreciated by a person of skill with the benefit of this description that the reflected infrared light from the infrared source 57c may be used to indicate the location and type of barrier, such as a solid wall, glass wall, or other barrier.

[0069] The processor 95c may include a central processing unit (CPU), a microcontroller, a microprocessor, a processing core, a field-programmable gate array (FPGA), an application-specific integrated circuit (ASIC), or similar. The processor 95c may cooperate with the memory storage unit 75c to execute various instructions stored thereon. In the present example, the memory storage unit 75c may store an operating system 430c that is executable by the processor 95c to provide general functionality to the device 50c, including functionality to identify and classify an area in a space. The processor 95c may also control the light source 55c with a light source controller 60c and process light data measured by the sensor 65c with an image processing engine 70c. In further examples, the memory storage unit 75c may be used to store additional applications that are executable by the processor 95c to provide specific functionality to the device 50c, such as functionality to control various components such as the sensor 65c, the communications interface 85c, and the light source 55c at the firmware level.

[0070] The processor 95c further operates an aggregator 80c to analyze tracking data generated by the image processing engine 70c. In addition, the processor 95c may also operate a grouping engine 90c grouping engine to associate the device 50c with a plurality of external devices autonomously. Accordingly, the device 50c may be controlled as part of a group of devices.

[0071] In the present example, the memory storage unit 75c may maintain databases to store various data used by the device 50c. For example, the memory storage unit 75c may include local tracking data 410c generated by the image processing engine 70c and external tracking data 420c received from proximate devices via the communications interface 85c. The memory storage unit 75c may additionally store an operating system 430c and additional instructions to carry out operations at the driver level as well as other hardware drivers to communicate with other components and peripheral devices, such as various user interfaces to receive input or provide output.

[0072] It should be recognized that features and aspects of the various examples provided above may be combined into further examples that also fall within the scope of the present disclosure.

Claims

What is claimed is:1 . A device comprising: a light source to emit light to illuminate a space; a light source controller to control the light source, wherein the light source controller is to change an intensity of the light emitted by the light source; a sensor to measure light data in the space, wherein the light data includes target information; an image processing engine to track a plurality of targets across a field of view, wherein the image processing engine generates local tracking data for each target of the plurality of targets; a memory storage unit to store the local tracking data for the plurality of targets; and an aggregator to analyze the local tracking data and to generate a classification of an area of the space, wherein the light source controller uses the classification to control the light source.

2. The device of claim 1 , further comprising a communications interface to communicate with an external device, wherein the communications interface is to receive external tracking data from the external device, and wherein the aggregator is to combine the local tracking data with the external tracking data to generate the classification of the device.

3. The device of claim 2, wherein the communications interface is to transmit the local tracking data to the external device.

4. The device of any one of claims 1 to 3, wherein the sensor is a low resolution sensor.

5. The device of any one of claims 1 to 4, further comprising a grouping engine to associate the device with a group of external devices autonomously, wherein the group covers the area with the classification.

6. The device of any one of claims 1 to 5, wherein the image processing engine detects changes in spatial signals in the field of view to identify each target of the plurality of targets.

7. The device of claim 6, wherein the local tracking data includes a path associated with each target of the plurality of targets across the field of view.

8. The device of claim 7, wherein the aggregator analyzes the path associated with each target of the plurality of targets to identify objects in the space, wherein the objects are used to generate the classification.

9. The device of any one of claims 1 to 8, wherein the aggregator generates a utilization map of the space.

10. A system comprising: a plurality of lighting devices disposed in a space, wherein each lighting device of the plurality of lighting devices comprises: a light source to emit light; a light source controller to control the light source, wherein the light source controller is to change an intensity of the light emitted by the light source;a sensor to measure light data in the space, wherein the light data includes target information; an image processing engine to track a plurality of targets across a field of view, wherein the image processing engine generates local tracking data for each target of the plurality of targets; a communications interface to communicate with other lighting devices of the plurality of lighting devices, wherein the communications interface is to receive external tracking data from the other lighting devices, and wherein the communications interface is to share the local tracking data with the other lighting devices; an aggregator to analyze the local tracking data and the external tracking data to group a subset of the plurality of lighting devices, wherein the subset is communicated to other devices; and a system controller to control the plurality of lighting devices, wherein the subset of the plurality of lighting devices is to be controlled together.11 . The system of claim 10, wherein each lighting device of the plurality of lighting devices further comprises a voting engine to obtain consensus of the subset.

12. The system of claim 10 or 11 , wherein each lighting device of the plurality of lighting devices further comprises a grouping engine to associate the lighting device with the subset of lighting devices autonomously, wherein each lighting device of the group covers an area with a common classification.

13. The system of any one of claims 10 to 12, wherein the image processing engine of each lighting device detects changes in spatial signals in the field of view of thelighting device to identify each target passing through the field of view of the lighting device.

14. The system of claim 13, wherein the local tracking data of each lighting device includes a path associated with each target passing through the field of view of the lighting device.

15. The system of claim 14, wherein the aggregator analyzes the path associated with each target to identify objects in the space, wherein the objects are used to group the subset of the plurality of lighting devices.

16. The system of any one of claims 10 to 15, wherein the aggregator of each lighting device generates a utilization map of the space.

17. The system of claim 16, wherein the system controller collects the utilization map from each lighting device of the plurality of lighting devices to generate an aggregated map.

18. A method comprising: measuring light data in a space with a sensor, wherein the light data includes target information; tracking a plurality of targets across a field of view of the sensor with an image processing engine; generating local tracking data for each target of the plurality of targets; analyzing the local tracking data to generate a classification of an area of the space; andcontrolling a light source to change an intensity of light emitted by the light source based on the classification.

19. The method of claim 18, further comprising: receiving external tracking data from an external device; and combining the local tracking data with the external tracking data to generate the classification of the area of the space, wherein analyzing the local tracking data comprises analyzing a combination of the local tracking data and the external tracking data.

20. The method of claim 19, further comprising transmitting the local tracking data to the external device.21 . The method of any one of claims 18 to 20, further comprising associating a lighting device with a group of external devices including the external device autonomously, wherein the group covers the area with the classification.

22. The method of claim 21 , wherein associating the lighting device comprises determining the angle between an aggregate path and an axis between an adjacent lighting device.

23. The method of any one of claims 18 to 22, further comprising detecting changes in spatial signals in the field of view to identify each target of the plurality of targets.

24. The method of claim 23, wherein the local tracking data includes a path associated with each target of the plurality of targets across the field of view.

25. The method of claim 24, wherein analyzing the local tracking data comprises analyzing the path associated with each target of the plurality of targets to identify objects in the space, and wherein the objects are used to generate the classification.

26. The method of any one of claims 18 to 25, further comprising generating a utilization map of the space.

Images