WO2019020395A1

WO2019020395A1 - Vcsel assembly

Info

Publication number: WO2019020395A1
Application number: PCT/EP2018/068963
Authority: WO
Inventors: Arjen Van Der Sijde; Rob ENGELEN; Nicola Bettina PFEFFER
Original assignee: Lumileds Holding BV; Lumileds LLC
Current assignee: Lumileds Holding BV; Lumileds LLC
Priority date: 2017-07-24
Filing date: 2018-07-12
Publication date: 2019-01-31
Anticipated expiration: 2020-01-24
Also published as: TW201909501A

Abstract

The invention describes a VCSEL assembly (1) comprising a plurality of individually addressable infrared-emitting VCSELs (10) arranged in a VCSEL array (10A); a controller (13) arranged to address individual VCSELs (10) of the VCSEL array (10A); and a plurality of primary optics elements (11a, 11b, 11c, 11o) arranged in a primary optics array (11A), wherein a primary optics element (11a, 11b, 11c, 11o) is shaped to spread the light (L10) originating from a VCSEL (10), and wherein the primary optics array (11A) is arranged over the VCSEL array (10A) such that each primary optics element (11a, 11b, 11c, 11o) corresponds to a VCSEL (10) of the VCSEL array (10A) and such that the emission axes (11a_X, 11b_X, 11c_X, 11o _X) of the primary optics elements (11a, 11b, 11c, 11o) diverge differently from a main axis (10A_X) of the VCSEL array (10A). The invention further describes a VCSEL module and a method of manufacturing a VCSEL assembly.

Description

VCSEL ASSEMBLY

FIELD OF THE INVENTION

The invention describes a VCSEL assembly, a VCSEL module, and a method of manufacturing a VCSEL assembly.

BACKGROUND OF THE INVENTION

VCSELs (vertical cavity surface emitting lasers) are being used in various applications such as 3D sensing, in which one or more infrared (IR) VCSELS are used to illuminate a "scene" or field. Here, the illumination source can be an array of VCSELs that share a common emission plane. In a 3D sensing application, light from the illumination source is directed at the target (e.g. a person, a face etc. in the scene or field that is being imaged), and the reflected light is detected, for example by a photodetector arrangement. However, the contents of a "scene" are effectively unpredictable, and a scene can include any number of objects that can be arranged at various distances from the illumination source. However, since illumination is constant over the emission plane, an object that is close to the illumination source may be overexposed, resulting in a loss of image information. An object that is far away from the illumination source may be underexposed, resulting in insufficient image information. Therefore, the known types of infrared 3D sensing systems can suffer from poor image quality.

Another application in which VCSELs may be used is that of optical wireless communication, commonly referred to as "LiFi". Visible or infrared light can be modulated to carry information that is transmitted from a light source to a photodetector. It is straightforward to set up a communication channel when the spatial relationship between light source and detector is fixed and constant. However, LiFi applications are being considered that may include mobile devices that communicate with the modules of a LiFi system installed in building, for example. In such an application, the mobile device must "discover" a LiFi module in order to be able to set up a communication channel, preferably without requiring the user to undertake any action. To this end, the light sources (e.g. infrared VCSELs) and photodetectors incorporated in the mobile device would have to face in various directions in order to discover a LiFi module, for example a ceiling-mounted LiFi module. However, this would require a very broad emission pattern which is energy consuming, and there would also be a risk of spurious reflections leading to a poor signal-to-noise ratio (SNR) or even data loss.

Therefore, it is an object of the invention to provide an improved VCSEL assembly.

SUMMARY OF THE INVENTION

The object of the invention is achieved by the segmented VCSEL assembly of claim 1; by the VCSEL module of claim 8; by the method of claim 13 of illuminating a scene; and by the method of claim 14 of manufacturing a segmented VCSEL assembly.

According to the invention, the VCSEL assembly comprises a plurality of individually addressable infrared-emitting VCSELs arranged in a VCSEL array; a controller arranged to address individual VCSELs of the VCSEL array; and a plurality of primary optics elements arranged in a primary optics array, wherein a primary optics element is shaped to spread the light originating from a VCSEL, and wherein the primary optics array is arranged over the VCSEL array such that each primary optics element corresponds to a VCSEL of the VCSEL array and such that the emission axes of the primary optics elements diverge differently from a main axis of the VCSEL array. Because the light sources are individually addressable, the inventive assembly may be referred to as a "segmented VCSEL assembly" in which the light sources act as "segments" of an overall light source.

In the context of the invention, the VCSEL assembly advantageously achieves directional emission, i.e. light emitted from the VCSEL array can be aimed in one or more directions by selectively activating the appropriate VCSEL(s). This is possible because the emission axes of the primary optics elements are different, and therefore diverge in different directions from the main axis of the VCSEL array. To simplify the discussion, a "VCSEL element" may be defined to comprise a VCSEL and its primary optics element. The beam of light leaving a VCSEL is emitted

perpendicularly from the emission surface of the VCSEL (i.e. along its "main emission axis") and will undergo refraction to some extent as it passes through the primary optics, so that the axis of the refracted light (referred to in the following as the

"refraction axis") leaving the primary optics will diverge from the main emission axis of the VCSEL. The main emission axis of a VCSEL may be defined as the axis perpendicular to the VCSEL surface, passing through the centre of the VCSEL.

Similarly, the main array axis of the VCSEL array may be defined as the axis perpendicular to the array substrate, passing through the centre of the VCSEL array.

In a preferred embodiment of the invention and in an extreme case, all refraction axes of the VCSEL elements of the VCSEL assembly are different, i.e. no two refraction axes (even if they lie in the same plane) are parallel. In a further preferred embodiment of the invention and in a less extreme case, groups of VCSELs elements may have parallel refraction axes and may be regarded as a cluster. For the sake of simplicity, any mention of a VCSEL in the following may be assumed to apply equally to a VCSEL cluster. In this way, the inventive VCSEL assembly makes it possible to redirect the light of every individual VCSEL or VCSEL cluster to a specific area of a total "field" or scene that is to be illuminated. The field or scene may be regarded as the region in space that can be illuminated by the VCSELs of the VCSEL array. The terms "field" and "scene" may be used interchangeably in the following. Since the VCSELs of the VCSEL array are individually addressable, individual

VCSELs can be selectively switched on and off as desired to distribute the VCSEL power selectively over the "field" or scene. The inventive VCSEL array therefore generates an "illuminance distribution" or "illumination field" in which the brightness or intensity of each individual VCSEL or VCSEL cluster is individually adjusted to the requirements of the corresponding region in the scene. One advantage of the inventive VCSEL assembly is that the divergence of any VCSEL beam is maintained, and may be widened even further to achieve a larger FWHM (full width at half maximum), while at the same time directing the light into different regions or sections of the field to be illuminated. This is in contrast to the known VCSEL assemblies that strive to minimise VCSEL beam divergence, for example to retain a FWHM in the order of 20°, and that involve complex

manufacturing techniques to individually orient dozens of VCSELs to achieve a wide illumination field. With the inventive VCSEL assembly, it is possible to illuminate a correspondingly large field or scene.

According to the invention, the VCSEL module comprises an embodiment of the inventive VCSEL assembly, arranged to direct light at a scene. The VCSEL module further comprises a photodetector arranged to detect light returned from the scene, and an analysis unit adapted to extract information from the returned light. In the context of the invention, the expression "returned light" can mean light that is reflected off an object in the scene and detected by the VCSEL module, or light that originates from a point in the scene and is detected by the VCSEL module. The nature of the returned light will depend on the application using the inventive VCSEL module. The controller is adapted to adjust the amount of infrared light emitted by each VCSEL on the basis of the returned light.

An advantage of the inventive VCSEL module is that it provides a way of regulating the illuminance of a target in the scene. For example, obtaining an accurate depth map in a time-of- flight (ToF) application is only possible when the quality of the detected signals is sufficient. In an imaging application, for example, a target that is too close to the VCSEL array will be overexposed so that the corresponding return signals are effectively "clipped" and have lost image information, while a target that is too far away will be underexposed so that the return signals are noisy. In the inventive VCSEL module, the controller is adapted to (iteratively if necessary) adjust the amount of infrared light emitted by a VCSEL on the basis of the signal strength between that

VCSEL and its corresponding target. For these reasons, the inventive VCSEL module may also be referred to as a "selective illumination module" in the following. The amount of light that is emitted by a VCSEL can be expressed in terms of power, i.e. a VCSEL of the inventive VCSEL module is powered to a specific intensity in order to achieve the desired light output to ensure that the quality of the return signal (reflected from a target) is sufficiently good to allow accurate phase detection. Apart from depth map computation, there are various other advantages of being able to selectively control the intensity of the light emitted by individual VCSELs in order to illuminate specific regions of the scene, as will be explained below. In one embodiment, the inventive infrared VCSEL module is realised for use in a 3D sensing application, for which the targets comprise regions in a field or "scene". The illumination field will illuminate the targets in the scene, and the 3D sensing application makes an image of this scene. Other imaging applications, such as face recognition applications for example, could benefit from the inventive approach. Each VCSEL element of the VCSEL array is provided as described above to illuminate a specific region of a scene by directing a beam of infrared light at that region. From the point of view of the VCSEL array, targets in the scene may be close up, in the middle distance, far away, etc. According to the invention, the method of illuminating a scene comprises the steps of providing an embodiment of the inventive VCSEL module; individually activating each VCSEL to direct light at a corresponding region of the scene; detecting a return light beam originating from each region of the scene; and iteratively adjusting the intensity of the light emitted by each VCSEL on the basis of the intensity of the return light beam.

The inventive method provides an advantageously straightforward way of "customizing" the illumination of a scene. Instead of simply illuminating all regions of a scene using a homogeneous illumination field in which all VCSELs would be driven at the same intensity, the inventive method provides a way of determining the intensity required for each VCSEL in order to optimally illuminate the corresponding scene region.

According to the invention, the method of manufacturing a VCSEL assembly comprises the steps of providing a plurality of individually addressable VCSELs in a VCSEL array; providing a controller to individually address the VCSELs of the VCSEL array; and arranging a plurality of primary optics elements over the VCSEL array such that each primary optics element corresponds to a VCSEL of the VCSEL array and such that the emission axis of a primary optics element diverges from the emission axis of the corresponding VCSEL and such that the emission axes of the primary optics elements diverge from a main axis of the VCSEL array.

A VCSEL array can be rectangular, comprising any suitable number of rows and columns, or can be hexagonal, triangular, etc. The plurality of individually addressable VCSELs can be achieved by means of a suitably organized matrix of contacts. The individual addressability of the VCSELs can for example be achieved by appropriate wafer- fab metal deposition using an appropriate crossed layout pattern for an array of contacts separated by an appropriated electrical insulation layer with contacting vias. For a high resolution VCSEL matrix with many VCSELs, layout and deposition techniques could be similar to those used commonly in the semiconductor industry, as in the design and manufacturing of displays.

The dependent claims and the following description disclose particularly advantageous embodiments and features of the invention. Features of the embodiments may be combined as appropriate. Features described in the context of one claim category can apply equally to another claim category.

A VCSEL array may be manufactured to comprise VCSELs that have different parameters or characteristics. However, such "customization" adds to the manufacturing costs. In the following, it may be assumed that the VCSEL array is advantageously made of essentially identical VCSELs arranged on a common substrate, and that the emission axis of each VCSEL of the VCSEL array is essentially perpendicular to the substrate of the VCSEL array.

Usually, as mentioned above, a VCSEL has a slightly divergent beam with a FWHM in the order of 20°. Most prior art VCSEL devices aim to conserve the narrowness of the beam, and generally also incorporate optical elements that reduce the beam divergence even further. In contrast, the primary optics elements of the inventive VCSEL assembly are specifically shaped to further spread the beam of light to achieve a FWHM of at least 40°, more preferably at least 50°, most preferably at least 60°. In this way, a scene can be optimally illuminated without requiring further secondary optics.

In the context of the invention, the VCSEL array and the primary optics array may be regarded as a single unit. The primary optics array may be understood to be permanently applied to the VCSEL array, for example by using a photolithographic technique as will be explained below. The primary optics array is essentially physically inseparable from the VCSEL array. An advantage of making the primary optics elements part of the VCSEL array is the attendant increase in safety, since it will be extremely difficult to remove the broadening optics without damaging the underlying VCSELs. With the inventive VCSEL assembly, the total VCSEL power of a VCSEL element will only ever be emitted in a deliberately broadened beam.

As mentioned above, a section or region of the total illumination field can be "covered" by a single VCSEL, and the VCSEL assembly can be designed so that the regions of the illumination field are distinct, i.e. with little or no overlap. Alternatively, the VCSEL assembly can be designed so that the regions of the illumination field overlap to a certain extent.

A primary optics element may be understood to be formed very close to the corresponding VCSEL, and may be formed essentially directly on top of the VCSEL. In a preferred embodiment of the invention, a primary optics element has the form of a facet, and the primary optics array comprises an array of differently shaped facets or facet elements. In other words, adjacent facets are of different sizes. For example, in one embodiment of the invention, the facets increase in size with increasing distance from the centre of the VCSEL array. Preferably, the facets are realised as dielectric facets. In a further preferred embodiment of the invention, the shapes and sizes of the facets are chosen such that the overall shape of the primary optics array follows a concave shape. Equally, the overall shape of the primary optics array may follow a convex shape. The facets may be realized as physically separate entities, or a single primary optics structure may be formed to comprise the plurality of functionally distinct primary optics elements.

Instead of differently-shaped triangular facets shaped to refract the light of the VCSELs by different amounts, in an alternative embodiment the primary optics array may comprise an array of identical lens elements. Each lens element is arranged to correspond to a single VCSEL of the VCSEL array. In such an embodiment, the different emission axis directions are achieved by offsetting the lens to different extents from the emission axes of the VCSELs. For example, the degree of offset may increase with increasing distance from the centre of the VCSEL array, and the direction of offset may depend on the position of the lens element relative to the centre of the array. This will be explained later with the aid of the diagrams.

The array of primary optics elements may be applied onto the VCSEL array using any suitable technique, for example photolithography, moulding, wet etching, replication etc. When photolithography is used, the primary optics elements may be formed as photonic crystal structures.

Depending on the array dimensions and overall shape, a VCSEL of the VCSEL array may be uncovered by a primary optics, i.e. the emission axis of such a VCSEL will remain unchanged. For example, this may apply to the central or innermost VCSEL in a hexagonal array, the innermost VCSEL(s) of a rectangular VCSEL array, a VCSEL arranged along the side of the VCSEL array, etc. In one approach as indicated above, differently-shaped primary optics elements can be provided for the VCSELs of the VCSEL array. This may result in a design in which - to use a simple example - a six-by-six VCSEL array requires nine different facet shapes, and four different ways of placing each facet ("north", "south", "east", "west"). In an alternative approach, the VCSEL array may comprise VCSELs with two or more different laser modes. The emission direction of a VCSEL element is then no longer only determined by facet shape, instead the emission direction is determined by the laser mode as well as facet shape. By appropriately distributing the laser modes over the VCSEL array, it may be possible to significantly reduce the number of different facet shapes that are required to achieve the desired variety in emission axis directions.

A relatively shallow facet will refract the light beam so that the refraction axis of the light beam of the VCSEL element is refracted slightly away from the VCSEL emission axis. In a preferred embodiment of the invention, more steeply inclined facets may be provided for certain VCSELs. The steeper inclination of the facet surface will result in total internal reflection (TIR) and a more severe re-direction of the VCSEL beam. In such an embodiment, the refraction axis of the facet with TIR may even cross the main emission axis of the array.

In one embodiment, the inventive infrared VCSEL module is realised for use in a 3D sensing application or similar imaging application, a face recognition application, etc. Here, the targets in the scene can be people, faces, objects etc. Each VCSEL element of the VCSEL array is provided as described above to illuminate a specific region of a scene by directing a beam of infrared light at that region. From the point of view of the VCSEL array, targets in the scene may be close up, in the middle distance, far away, etc.

As mentioned already, overexposure occurs in an imaging application when a target receives too much light from its corresponding illumination source. The corresponding image region will not contain sufficient image information due to the detector signal (of the camera sensor) being saturated. Underexposure occurs when a target does not receive sufficient light, and the corresponding image region will be noisy. For a time-of- flight unit that measures distances to the targets, the poor signal quality is problematic as it compromises the final result by erroneous distances and results in areas in the scene for which distance cannot be measured. This problem is overcome by the inventive VCSEL module, since the intensity of each VCSEL can be regulated independently, thereby allowing each corresponding region of the scene to be illuminated correctly. With the inventive method, the signals arriving at the

photodetector are always "informative", i.e. none of the signals are clipped owing to overexposure, or noisy owing to underexposure. Because of the favourable SNR that is obtained once the correct brightness is set over the entire scene, an accurate depth map can be generated, for example in an image-sensing application. A depth map can be obtained for example by measuring the distances travelled between the VCSELs and a photodetector, or by phase detection in which the light from the emitter (i.e. the VCSEL) is modulated and the detector examines the phase shift between emitted and reflected light.

By controlling each VCSEL element appropriately, the intensity or brightness of each target, i.e. each region in the scene, can be precisely controlled. In this way, "overexposed" and "underexposed" image regions can be avoided. The controller can appropriately reduce the amount of light directed at a close object.

Similarly, the controller can increase the intensity of the light directed at a remote object (within the capability of the emitter). Since the individual VCSELs can be addressed and regulated individually, the light power can selectively be distributed over the "scene". The selective intensity increase will not result in overexposing other areas of the scene, for example adjacent targets in the scene, as would be the case with a conventional non-addressable VCSEL array. The depth map can thereby be generated for the complete scene at the same time by using a time-of- flight technique, "light radar" (LIDAR), structured light image processing technique, or any other suitable approach. To this end, the distance estimation module can comprise an array of photodetectors, preferably the photodetector array of the camera sensor. In such an embodiment, each photodetector or pixel area of the camera is associated with a VCSEL element of the VCSEL module in a one-on-one arrangement. Alternatively, the segmented VCSEL could also be used in combination with a single photodetector. In such an embodiment, the scene can be scanned by sequentially switching the VCSELs "on" and then "off, one after the other. As explained above, during the "on" time of a VCSEL, its intensity can be tuned adaptively or iteratively to obtain a good signal-to- noise ratio for the corresponding region in the scene. On the basis of the signal strength, the distance estimation module will determine the distance between the VCSEL and the corresponding region or target in the scene, and the direction to any object in that region in the scene is given by the combination of primary optics element and the VCSEL that is currently switched "on".

In a further embodiment, the inventive infrared VCSEL module is realised for use in an optical wireless communication system or "LiFi" system. A LiFi system may be realised to comprise a number of light sources that can transmit information in the form of modulated light, and a number of detectors or receivers to detect modulated light. In such an embodiment of the inventive VCSEL module, each VCSEL of the VCSEL array is arranged to emit a beam of infrared light, and may be detected by a LiFi receiver within a "target" in the field of the VCSEL array. A LiFi receiver may comprise some suitable kind of detector for detecting a modulated light signal, as will be known to the skilled person.

In such an embodiment, the best communication direction between a VCSEL and a LiFi receiver may be determined on the basis of the signal strength as a function of the beam direction of infrared light emitted by that VCSEL. This could be done by the controller, which may sequentially activate and de-activate the VCSELs one by one, and which may then select the VCSEL(s) that resulted in the best or strongest return signal from a LiFi transmitter. In this way, only those VCSELs are activated that are directed towards a LiFi receiver and within reasonable distance of a LiFi receiver. One of the useful features of a LiFi communication system is that multiple communication channels can be maintained from a single device - for example a mobile device with an embodiment of the inventive VCSEL module - in different directions and at the same time.

For example, LiFi may be used in a large building such as a conference area, hospital, concert arena, airport etc., illuminated by many lights and visited by varying numbers of people. Most of the visitors may be assumed to carry a smartphone or similar mobile device that incorporates an embodiment of the inventive VCSEL module. Alternatively, an embodiment of the inventive VCSEL module may be issued to each visitor as a visitor's badge or lanyard, for example. Multiple LiFi receivers can be distributed within the large building, for example each LiFi transmitter (i.e. lamp) on the ceiling may be provided with one or more LiFi receivers. As a visitor moves about in the building, VCSELs of the VCSEL array are selectively activated and deactivated so that only the VCSEL(s) within range of the nearest detector(s) are active. The signal strength between VCSELs and targets are updated at regular intervals, and the controller of the VCSEL module will continually analyse the signal strength information to selectively activate/deactivate VCSELs as appropriate.

The LiFi system can use the information received by multiple such mobile devices, for example to save power by only turning on lamps as needed in the areas of the building where visitors are detected or by adjusting the illumination intensity.

Illumination in sparsely visited areas of a building, for example in rooms of a museum, can be turned on (or increased) when a visitor approaches and turned off again (or dimmed) as the visitor departs. By means of the communication channel between the LiFi system and the visitors' mobile devices, it is possible to track the movements of visitors in the building and for lighting to be precisely tailored to the momentary requirements, thus saving energy. Similarly, the sending power of the mobile device can be kept to a favourable minimum by only communicating with nearby LiFi receivers when the signal strength is acceptable. With an appropriate feedback signal for the user, privacy can be ensured by allowing the user to communicate only with LiFi receivers of approved networks. Other objects and features of the present invention will become apparent from the following detailed descriptions considered in conjunction with the

accompanying drawings. It is to be understood, however, that the drawings are designed solely for the purposes of illustration and not as a definition of the limits of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig 1 shows an embodiment of the inventive VCSEL assembly;

Fig 2 shows an alternative realisation of the embodiment of Fig 1; Fig 3 illustrates a relationship between a VCSEL array and an

illumination field;

Fig 4 shows a block diagram of an embodiment of the inventive

VCSEL module;

Fig 5 illustrates a 3D sensing application using an embodiment of the inventive VCSEL module;

Fig 6 illustrates an optical wireless communication application using an embodiment of the inventive VCSEL module;

Figs 7 and 8 show alternative embodiments of the inventive VCSEL

assembly.

In the drawings, like numbers refer to like objects throughout. Objects in the diagrams are not necessarily drawn to scale.

DETAILED DESCRIPTION OF THE EMBODIMENTS

Fig 1 shows a simplified illustration of a VCSEL assembly 1 according to an embodiment of the invention. Here, the VCSEL array 10A is shown to comprise seven rows of VCSELs 10, and the cross-section is taken through the centre of the array 10A. The innermost VCSEL 10 is not covered by a primary optics element, and the beam L₁₀ emitted by the central VCSEL 10 travels in the same direction as the main array axis 10A X of the VCSEL array 10A. The remaining VCSELs 10 are covered by differently shaped primary optics elements in the form of facet-shaped bodies 1 la, 1 lb, 11c. The effect of the different facet shapes is to refract the VCSEL light beams L₁₀ from those VCSELs 10 to different extents, and outward from the main axis 10 A X of the VCSEL substrate. Light from the VCSEL underneath facet 1 la is directed along the corresponding emission axis 1 la_X. The diagram shows two such facets 11a facing in opposite directions, resulting in two differently refracted light beams. Similarly, light from the VCSEL underneath facet 1 lb is directed along the corresponding emission axis 1 lb_X, while light from the VCSEL underneath facet 1 lc is directed along the corresponding emission axis 1 lc_X. Only the non-refracted beam from the innermost VCSEL travels in the same direction as the main axis 10A X of the VCSEL substrate. The diagram also illustrates the effect of the facets on the width of the refracted light beams. The otherwise narrow VCSEL beam (as shown by the beam leaving the central VCSEL) is broadened by the facets 11a, 1 lb, 1 lc to achieve a greater FWHM and to enable the VCSEL module 1 to illuminate a relatively large scene.

Fig 2 shows an alternative realisation of the embodiment of Fig 1. Here, instead of individual facet elements, the primary optics is realised as a single unit, but with facet regions that refract in the same way as the individual facets of Fig 1.

Fig 3 illustrates a possible relationship between a VCSEL array 10A (upper part of the diagram) and an illuminance distribution or illumination field F (lower part of the diagram). The illuminance distribution F will illuminate a scene. The diagram shows a plan view of an embodiment of the VCSEL assembly 10A, 11 A as explained in Fig 1. Here, the VCSEL array 10A is a 5 x 5 array. The position of each individual VCSEL 10 in the array can be specified by its row (Rl - R5) and column (CI - C5). In the same way, the illumination field F can be divided into field regions, whereby each field region will be illuminated by a VCSEL 10 of the VCSEL array 10A. In this example, the VCSEL in row R5 and column CI will illuminate field region F51; the VCSEL in row R4 and column C4 will illuminate field region F44, etc. By appropriate choice of primary optics shape, the field regions may overlap to a certain extent as shown in the diagram, or not at all. The diagram shows a very simplified example in which the VCSEL beams have the same FWHM. Of course, the shapes of the primary optics may be chosen to broaden the VCSEL beams by different extents.

Fig 4 is a very simplified block diagram of part of an embodiment of the inventive VCSEL module 3, indicating a relationship between a VCSEL array 10A, a photodetector 2 and an analysis unit 12, which in this case is realised as a distance measurement unit 12. The photodetector 2 can be the image sensor 2 of a camera of a mobile device incorporating the inventive VCSEL module 3, for example. For use in a known type of time-of- flight application, the image sensor will be split in multiple regions to realize the spatial resolution of the image. In the diagram, the pixel array of the photodetector 2 is shown in a very simplified manner, and it will be understood that a photodetector 2 could comprise thousands of pixels (e.g. when a ToF sensor is used) or millions of pixels (when a camera image sensor is used). To establish a depth map for a scene, each VCSEL 10 emits a beam of light (shown collectively as L₁₀ in the diagram), and the return beams (shown collectively as Ls in the diagram), are detected by the image sensor 2. Each return beam is a light beam that was reflected from a target, i.e. an object in the corresponding field region of the scene S. During the on-time of a VCSEL, the brightness or intensity of the corresponding return beam is sensed by the corresponding pixels of the photodetector 2 and is directly related to the extent to which that scene region was exposed. If the signal is weak, the illuminator signal in that region of the scene can be increased by increasing the VCSEL power; if the signal is clipped (overexposed) the illuminator VCSEL emitting in that region of the scene can be down-regulated or even switched off. The controller 13 selectively and iteratively activates/deactivates/regulates the individual VCSELs 10 as appropriate. With the optimized illumination, the information provided by the photodetector 2 can be analysed using known image processing techniques (for example time-of- flight) to determine a depth map FDM for the scene. The optimized illuminance distribution results in an optimized signal-to-noise ratio over the complete scene, avoiding over- exposed or under-exposed image regions. As mentioned above, the selective illumination module can use a simplified type of photodetector. For example, a single photodetector could be used and the illumination module can sequentially switch the VCSELs on and off to scan the image and to determine the extent of exposure of each image region.

Fig 5 illustrates a 3D sensing application. Here, the targets comprise objects in a scene S. With a prior art 3D sensing apparatus that illuminates all parts of the field evenly, targets close to the illumination source may be overexposed, as shown in the upper part of the diagram. Here, the person on the left-hand side of the scene S has been illuminated too strongly (i.e. overexposed) and image detail has been lost. The person in the background has been insufficiently illuminated, and insufficient image detail was obtained. In the inventive VCSEL module, each VCSEL element of the VCSEL assembly can be individually controlled to illuminate a specific region of the scene S. In this example, a 4 x 5 VCSEL array is being used, and the illuminance distribution F is virtually divided into 20 regions, so that the VCSEL in row Rl and column CI will illuminate field region Fl 1, the VCSEL in row R4 and column 4 will illuminate field region F42, etc. Each field region of the illuminance distribution F will illuminate a corresponding region in the scene S. Using the inventive VCSEL module, an image is taken and the brightness distribution is analysed. The controller 13 will selectively reduce the power to the three VCSELs in column C2 that correspond to illumination field regions F22, F32, F42 in order to reduce illumination of the corresponding scene regions (e.g. the person closer to the VCSEL module in the above example), and may also use information regarding the ambient lighting levels. Such information can be provided by photodiodes of the VCSEL assembly, for example. The controller 13 may also selectively adjust the power to the VCSELs in column C4 that correspond to illumination field regions F24, F34, F44 in order to optimise illumination of the corresponding scene regions, that is to increase the illuminance for the person further from the VCSEL module. The lower part of the diagram shows the same scene S, but with improved illumination for both people and a corresponding improvement in information content. The diagram has used only a limited number of image regions to explain the method, and it will be understood that the resolution of the VCSEL array and scene can be significantly greater. Using the above scene S as an example, it may be possible to adjust the illuminance over relatively small scene regions, e.g. regions as small as the arm and chest or eyes and nose of a person in the above scene.

Fig 6 illustrates a LiFi embodiment of the inventive VCSEL module. The diagram indicates a large building with visitors that move about freely. LiFi-enabled lamps LI, L2, L3 (LiFi "modules") illuminate various rooms or spaces in the building. LiFi "communications channels" are set up as the infrared-emitting VCSELs of the VCSEL assemblies in the visitors' mobile devices M emit light beams L₁₀ that are detected by receivers of the LiFi modules LI, L2, L3, and light from an emitter (e.g. LED lamps) of a LiFi module LI , L2, L3 is detected by the photodetector of the VCSEL module. Each room in the building with an arrangement of LiFi lamps LI, L2, L3 can therefore be regarded as a "scene" S from the point of view of the inventive VCSEL module, and any light beam from an emitter of a LiFi module LI, L2, L3 is a "return beam" from the point of view of a VCSEL module. The illumination system can be controlled to regulate the individual lamps LI, L2, L3 according to the number of visitors in the various spaces in the building. This can save energy and reduce the running costs for the building. In a mobile device, the controller of the VCSEL assembly can activate and deactivate the VCSELs accordingly as the visitor moves about, to only activate the VCSEL that is directed towards the nearest LiFi receiver. Another use of such a LiFi embodiment might be for the continual assessment of evacuation routes from a building depending on the concentration of visitors in the various regions of the building. Another use of such a LiFi embodiment might be to send the user information specific to his position in the building (information about a museum exhibit, price information for a product on display, etc.). Such information can be sent by a module LI, L2, L3 of the LiFi system and detected by an image sensor in the user's mobile device, for example, once a LiFi communication channel has been set up between the VCSEL module in the mobile device and a module LI, L2, L3 of the LiFi system. In the case of more sensitive information, such information can be requested by the user when close to the LiFi emitter of a module LI, L2, L3 and when the communication channel has been established.

Fig 7 shows an alternative embodiment of a VCSEL assembly using facets to refract the light from the VCSELs 10. In this embodiment, which is otherwise similar to that of Fig 1, the outermost facets 11c are shaped to refract the light by total internal reflection (TIR) along emission axes 1 lc_X, resulting in those beams of light "crossing over" the main axis 10A X of the array 10A. .

Fig 8 shows an alternative embodiment of a VCSEL assembly. Here, the primary optics elements 1 lo are realised as lenses having the same size and form. To achieve the desired directional emission of the light beams, i.e. will essentially all beams (expect the central beam) diverging away from the main array axis 10A X, the primary optics elements 1 lo are offset so that, with increasing distance from the array centre, the offset increases between a central axis 1 I X of a primary optics element 1 lo and the main emission axis of the VCSEL 10. The main array axis 10A X of the VCSEL array 10A may be defined as the axis perpendicular to the array substrate, passing through the centre of the VCSEL array 10A. Similarly, the main emission axis of a VCSEL 10 may be defined as the axis perpendicular to the VCSEL surface, passing through the centre of the VCSEL. Although the present invention has been disclosed in the form of preferred embodiments and variations thereon, it will be understood that numerous additional modifications and variations could be made thereto without departing from the scope of the invention. For example, the inventive VCSEL module with its ability to selectively illuminate a field can be used in face recognition or similar image processing applications which use infrared illumination and which may benefit from a selective illumination module. Another application for the inventive VCSEL assembly may be in communication between moving objects such as cars or drones, in applications which may benefit from the inherent selective information exchange that is made possible by the inventive VCSEL module. For example, in a car-to-car communication application, a signal from a car could be targeted only to any other cars travelling in the lane into which that car intends to move. Furthermore, instead of using VCSELs as light sources, a segmented assembly that makes use of other semiconductor light sources such as LEDs is also possible. For the sake of clarity, it is to be understood that the use of "a" or "an" throughout this application does not exclude a plurality, and "comprising" does not exclude other steps or elements. The mention of a "unit" or a "module" does not preclude the use of more than one unit or module. REFERENCE SIGNS:

VCSEL assembly

VCSELs 10

emission axis 10_X

VCSEL array 10A

main axis 10A X

primary optic elements 11a, l ib, 11c, l lo

primary optic array 11A

refraction axis 1 la_X, 1 lb_X, 1 lc_X, 1 lo _X distance estimation module 12

controller 13

VCSEL apparatus 14

image sensor 2

photodetector 20

VCSEL module 3 emitted light L 10

return light Ls

scene S

illuminance distribution F

region in illumination field F₅₁, F₄4

LiFi module L1, L2, L3

Claims

CLAIMS:

1. A VCSEL assembly (1) comprising

- a plurality of individually addressable infrared-emitting VCSELs (10) arranged in a VCSEL array (10A);

a controller (13) arranged to address individual VCSELs (10) of the VCSEL array (10A); and

a plurality of primary optics elements (11a, 1 lb, 11c, 1 lo) arranged in a primary optics array (11A), wherein a primary optics element (11a, 1 lb, 11c, 1 lo) is shaped to spread the light (L₁₀) originating from a VCSEL (10), and wherein the primary optics array (11 A) is arranged over the VCSEL array (10A) such that each primary optics element (11a, 1 lb, 11c, 1 lo) corresponds to a VCSEL (10) of the VCSEL array (10A) and such that the emission axes (1 la_X, 1 lb_X, 1 lc_X, 1 lo X) of the primary optics elements (11a, 1 lb, 11c, 1 lo) diverge differently from a main axis

(10A X) of the VCSEL array (10A).

2. A VCSEL assembly according to claim 1, wherein the primary optics array (11A) comprises an array of differently shaped facet elements (11a, 1 lb, 11c).

3. A VCSEL assembly according to claim 2, wherein the overall shape of the primary optics array (11 A) follows a convex shape.

4. A VCSEL assembly according to claim 1, wherein the primary optics array (11A) comprises an array of identical lens elements (Ho), and wherein the lens element (1 lo) of a VCSEL (10) is offset from the emission axis (10_X) of that VCSEL (10).

5. A VCSEL assembly according to any of the preceding claims, wherein the emission axis (10 X) of each VCSEL (10) of the VCSEL array (10A) is essentially perpendicular to a VCSEL array substrate.

6. A VCSEL assembly according to any of the preceding claims, wherein a primary optics element (11a, 1 lb, 11c, 1 lo) is shaped to achieve a FWHM of at least 40°, more preferably to achieve a FWHM of at least 50°, most preferably to achieve a FWHM of at least 60°.

7. A VCSEL assembly according to any of the preceding claims, wherein the VCSEL array (10A) comprises VCSELs (10) with different laser modes.

8. A VCSEL module (3) comprising

- a VCSEL assembly (1) according to any of claims 1 to 7, arranged to direct light at a scene (S);

a photodetector (2) arranged to detect light (Ls) returned from the scene (S); and an analysis unit (12) adapted to extract information from the returned light (Ls); wherein the controller (13) is adapted to adjust the intensity of the light emitted by a VCSEL (10) on the basis of the intensity of the returned light (Ls).

9. A VCSEL module according to claim 8, realised for use in a 3D sensing system, wherein the targets comprise objects in a scene (S), and each VCSEL (10) of the VCSEL array (10A) is arranged to illuminate a specific region of the scene (S).

10. A VCSEL module according to claim 9, wherein the analysis unit (12) comprises a distance estimation module (12) realised for use in any of a time-of- flight camera arrangement or a structured light 3D scanner arrangement.

11. A VCSEL module according to claim 8, realised for use in an optical wireless communication system, wherein the targets comprise modules (LI, L2, L3) of the optical wireless communication system, and wherein a VCSEL (10) of the VCSEL array (10A) is arranged to emit a beam of infrared light (L₁₀) that can be detected by a module (LI, L2, L3) of the optical wireless communication system.

12. A VCSEL module according to any of claims 8 to claim 11, wherein the photodetector (2) comprises an image sensor (2) of a camera.

13. A method of illuminating a scene (S), comprising the steps of

providing a VCSEL module (3) according to any of claims 8 to 12;

individually activating each VCSEL (10) to illuminate a corresponding region of the scene (S);

detecting return signals (Ls) correlated to the regions of the scene (S); and

- iteratively adjusting the intensity of the light emitted by each VCSEL (10) on the basis of the return signals (Ls).

14. A method of manufacturing a VCSEL assembly (1), comprising the steps of

providing a plurality of individually addressable VCSELs (10) in a VCSEL array (1 OA);

providing a controller (13) to individually address the VCSELs (19) of the VCSEL array (1 OA);

arranging a plurality of primary optics elements (11a, 1 lb, 11c, 1 lo) in a primary optics array (11A), wherein a primary optics element (11a, 1 lb, 11c, 1 lo) is shaped to spread the light (L₁₀) originating from a VCSEL (10), and arranging the primary optics array (11A) over the VCSEL array (10A) such that each primary optics element (11a, 1 lb, 11c, 1 lo) corresponds to a VCSEL (10) of the VCSEL array (10A) and such that the emission axes (1 la_X, 1 lb_X, 1 lc_X, l lo X) of the primary optics elements (11a, 1 lb, 11c, 1 lo) diverge differently from a main axis (lOA X) ofthe VCSEL array (10A).

15. A method according to claim 14, wherein the plurality of primary optics elements (11a, 1 lb, 11c, 1 lo) is provided in any of a photolithography step, a moulding step, a wet etching step, a replication step.