阅读说明：本技术 提供结构光的衍射光学设备 (Diffractive optical device providing structured light ) 是由 T.R.M.塞尔斯 S.H.查克玛简 G.M.莫里斯 于 2019-06-28 设计创作，主要内容包括：一种衍射光学元件,包括沿着光学材料的表面的微结构,微结构具有相位轮廓,以将输入照射衍射成多个不同衍射级的结构光；其中,相位轮廓至少部分相位解缠。还公开了生成衍射光学元件的方法。(A diffractive optical element comprising a microstructure along a surface of an optical material, the microstructure having a phase profile to diffract input illumination into structured light of a plurality of different diffraction orders; wherein the phase profile is at least partially phase unwrapped. Methods of producing diffractive optical elements are also disclosed.)

1. A diffractive optical element comprising:

a microstructure along a surface of an optical material, the microstructure having a phase profile to diffract input illumination into structured light of a plurality of different diffraction orders;

wherein the phase profile is at least partially phase unwrapped.

2. The diffractive optical element as claimed in claim 1, wherein said phase profile is

3. The diffractive optical element as claimed in claim 1, wherein said phase profile is below full unwrapping.

4. The diffractive optical element as claimed in claim 1, wherein said optical material is a single optical material.

5. The diffractive optical element as claimed in claim 4, wherein a surface of said single optical material is a first surface opposite a second surface of said single optical material.

6. The diffractive optical element as claimed in claim 5, wherein said second surface has a microstructure identical to the microstructure along the first surface of said single optical material.

7. The diffractive optical element as claimed in claim 5, wherein said second surface has a microstructure that is different from the microstructure along the first surface of said single optical material.

8. The diffractive optical element as claimed in claim 1, wherein said optical material is a composite optical material comprising two or more different optical materials.

9. The diffractive optical element as claimed in claim 8, wherein the composite has a first outer surface opposite a second outer surface.

10. The diffractive optical element as claimed in claim 9, wherein said second outer surface has a microstructure that is the same as or different from the microstructure along the first outer surface of said composite optical material.

Technical Field

The present invention relates to a diffractive optical element comprising a microstructure along a surface of an optical material, the microstructure having a phase profile to diffract input illumination into structured light of a plurality of different diffraction orders, wherein the phase profile is at least partially phase unwrapped. In one aspect, the structured light does not include a hot spot of zero diffraction order. Structured light, on the other hand, does not comprise any artifacts that substantially affect the light intensity uniformity of the structured light provided at each diffraction order. Diffractive optical elements may be used in illumination systems, such as to provide structured light for three-dimensional (3D) sensing, beam shaping, and display.

Background

In various applications, a radiation source consisting of a single beam must be converted into a plurality of spatially separated sub-beams (beams). Each beamlet propagates to a predetermined direction at a predetermined power with respect to other beamlets within the plurality of beamlets. This type of device is commonly referred to as a "beam splitter", which relies on the optical phenomenon of diffraction. Each sub-beam generated by the beam splitter is of one diffraction order. Each diffraction order is associated with a fraction of its efficiency or emission power. Known methods of implementing such solutions are based on Diffractive Optical Elements (DOEs). Diffractive optical elements can allow for the projection of image patterns by patterning a surface relief (relief) on a substrate material, such as fused silica or polymer. The surface relief acts on the incident illumination source by changing its wavefront phase content so that the diffraction pattern observed in the far field is tailored to the desired format. An example of an illumination system is shown in fig. 1. An illumination source, typically a laser of dominant wavelength lambda, provides an illumination beam to be projected. Depending on the optical requirements of the optical arrangement and system, collimating optics may be used to collimate the illumination beam. The DOE may act on the illumination beam and may modify the illumination beam such that the observed diffraction pattern projects some specific pattern. In the example of fig. 1, the illumination source is transformed into 5 spatially separated beams (diffraction orders) in the far field. In general, the illumination beam transformation may follow arbitrary specifications, such as spot arrays or more complex images. The DOE itself can be described as a grating with generally complex unit cells. An example of a typical DOE is shown in fig. 2. In particular, the DOE may include unit cells (represented by dashed lines) that may be repeated in a pattern or array, such as a 3 x 3 array.

The DOE may be a thin element with a phase depth of pi (in the case of a binary element) or 2 pi (in the case of a continuous profile element). The DOE may be limited to single wavelength operation, which is commonly referred to as the "design wavelength". When deviating from the design wavelength, undesirable effects may occur. For example, the relative power of the zero diffraction order tends to increase rapidly compared to other diffraction orders. This important effect is unavoidable. In some cases, the zero diffraction order may be physically blocked. While a strong zero order may not pose a problem in some cases, its presence is absolutely unacceptable in certain applications. One such application is 3D sensing, where a DOE is used to project a desired pattern of structured light, which is a particular distribution of diffraction orders with a specified position and relative intensity. In 3D sensing applications, the structured light pattern may be projected directly to the viewer with an infrared laser. In this case, a strong zero order directed to the viewer may constitute an eye safety issue and is therefore unacceptable for this type of application.

To manage the zero diffraction order problem while still using DOEs, existing approaches typically require sacrificing efficiency to produce a useful solution. In one example of such an approach, double-sided DOEs have been proposed, such as described in us patent No.8,630,039. In this method, two DOEs are fabricated on opposite sides of a substrate. The first DOE creates a pattern covering a narrower field of view with a more complex spot distribution. The second DOE, with a wider field of view, produces a simpler spot distribution. The combination of two DOEs enables a solution to be available that addresses the aforementioned eye safety issues. However, typical embodiments of both DOE approaches are inefficient because each DOE has a binary phase function with a maximum theoretical efficiency of 80% (excluding surface losses). Thus, the combination of two DOEs provides the theoretical maximum of 64% efficiency. The actual efficiency is more likely to be around 50% if surface losses are taken into account.

It is therefore desirable to provide an optical diffraction apparatus using a single facet for efficiently generating a structured light distribution with predetermined diffraction order positions while avoiding hot spots of zero diffraction orders.

Disclosure of Invention

In one aspect, disclosed herein is a diffractive optical element comprising a microstructure along a surface of an optical material, the microstructure having a phase profile to diffract input illumination into structured light of a plurality of different diffraction orders; wherein the phase profile is at least partially phase unwrapped.

In another aspect, disclosed herein is a method for generating a diffractive optical element, comprising: calculating a phase profile that produces the structured light pattern; unwrapping the calculated phase profile of the structured light pattern to obtain a 2 π P unwrapped phase profile; generating a 2 pi P unwrapped phase profile; based on the generated 2 π P unwrapped phase profile, microstructures are fabricated along the surface of the optical material.

Drawings

The foregoing objects, features and advantages of the invention will become more apparent from the following description when read in conjunction with the accompanying drawings, in which:

FIG. 1 is an optical diagram illustrating an illumination system including an illumination source that provides illumination energy to a diffractive optical element via collimating optics;

FIG. 2 is a diagram of the two-dimensional diffractive optical element of FIG. 1;

FIG. 3 is a graphical representation of structured light in the form of a 3 × 3 array of light spots, showing zero diffraction orders from the diffractive optical element of FIG. 1;

fig. 4 is a block diagram of a diffractive optical element according to the present invention.

Fig. 5A is an exploded block diagram of a diffractive optical element according to another aspect of the present invention.

FIG. 5B is a block diagram of the diffractive optical element of FIG. 5A;

fig. 6 is a block diagram of a diffractive optical element according to another aspect of the present invention.

FIG. 7 is a graph showing the ratio ρ of binary diffractive elements projecting 100 spots, where the phase retardation is 0 and 9 π radians;

8A-8E illustrate five graphs showing quantization of the phase function for the original 2, 4, 8, and 16 levels;

FIG. 9 is a graph showing a normalized ratio ρ/Ns as the current N_e＝N₀A function of the wavelength detuning (α) of the pi-diffraction element with pi phase shift at + 1;

FIG. 10 is a graph showing the 2 π -wrap of the diffractive lens phase profile (solid line) from a typical lens phase profile (dashed line);

FIG. 11 is a phase profile in an example of a one-dimensional structured-light phase profile, which generates 41 spots or sub-beams of structured light;

FIG. 12 is a graph of a one-dimensional diffraction pattern resulting from the ideal phase profile of FIG. 10, showing-20 to +20 diffraction orders that generate 41 spots or beamlets;

FIG. 13 is a graph of the diffraction pattern produced by the phase profile of FIG. 11 with a 5% depth error, showing the higher intensity (or power) of the zero diffraction order compared to the other orders in terms of output illumination efficiency;

FIG. 14 is a graph of a phase profile (solid line) in the form of an 8 π -wound phase profile of the initial phase profile of FIG. 11 when fully unwound, wherein the initial phase profile of FIG. 11 (dashed line) is shown to illustrate a reduction of phase discontinuity in the 8 π -wound phase profile;

FIG. 15 is a graph of the diffraction pattern produced by the 8 π wound phase profile of FIG. 14 with a 5% depth error, showing an excessive uniformity intensity error among 41 diffraction orders;

FIG. 16 is a graph of a diffraction pattern in the form of a 4 π wound phase profile with a 5% depth error of the initial phase profile of FIG. 11, when partially unwrapped, showing reduced uniformity intensity error among 41 diffraction orders and avoided zero diffraction order hot spots;

FIG. 17 is a graph of a phase profile showing a 4 π wound phase profile (solid line) with a 5% depth error resulting in the diffraction pattern of FIG. 16 when the phase profile is partially unwrapped, wherein the initial phase profile of FIG. 11 (dashed line) is shown to show a reduction of phase discontinuity in the 4 π wound phase profile; and

FIG. 18 is a flow chart of a method for generating a diffractive optical element that provides structured light in accordance with the present invention.

Detailed Description

The diffractive optical element 10 of the present invention can address the disadvantages of conventional diffractive optical elements and periodic microlens arrays by including microstructures 11 along the surface of the optical material 12, the microstructures 11 having a phase profile to diffract input illumination into structured light of a plurality of diffraction orders, wherein the phase profile is at least partially phase unwrapped. In one aspect, the phase profile of the microstructure is a 2 π P unwrapped phase profile. The diffractive optical element 10 of the present invention does not exhibit the deleterious presence of high intensity zero diffraction orders that result in hot spots in the far field. In addition, the diffractive optical element 10 can operate without a high intensity zero diffraction order over a wider wavelength range than conventional diffractive optical elements.

As shown in fig. 4-6, the diffractive optical element 10 can include microstructures 11 along the surface of the optical material 12. The optical material 12 may be any material capable of manipulating energy flow (such as light or electromagnetic radiation) in the ultraviolet, visible, and infrared spectral regions. The optical material 12 may be selected based on the properties of the material, such as transparency, transmittance, refractive index, etc. Non-limiting examples of the optical material 12 include plastic (e.g., polymer), glass, or silica. The optical material 12 may be a single optical material (fig. 4 and 6), or may be a composite optical material comprising two or more different optical materials (fig. 5).

As shown in fig. 4-6, the diffractive optical element 10 can include an optical material 12 and microstructures 11 along a surface of the optical material 12. The microstructures 11 may be formed using any conventional technique, such as hot embossing, injection molding, reactive ion etching, or ion beam milling. The microstructures 11 may have a phase profile, such as a 2 π P unwrapped phase profile, that diffracts the input illumination into structured light of a plurality of different diffraction orders. The phase profile may be at least partially phase unwrapped along the surface 14a of the optical material 12.

With respect to fig. 4, the diffractive optical element 10 may include a first surface 14a, and the first surface 14a may extend along one or two dimensions, each dimension perpendicular to the depth of the phase profile extension. The diffractive optical element 10 may comprise a single optical material 12. Microstructures 11 may be present along first surface 14a of optical material 12. In particular, the microstructures 11 may be formed along the surface 14a of the optical material. The surface of the single optical material 12 may be a first surface 14a, and the first surface 14a may be opposite a second surface 14b of the single optical material 12. In one aspect, the second surface 14b of the optical material 12 may be absent of microstructures 11, i.e., the second surface 14b is flat.

In operation, the diffractive optical element 10 can receive input illumination along the first surface 14a through the microstructures 11. The input illumination may be any light source, e.g. a beam of light from a coherent light source, such as a laser, which is used for 3D sensing of light having infrared wavelength(s) or range.

The structured light of the plurality of different diffraction orders may have various shapes, forms and/or patterns. Non-limiting examples of structured light include spot arrays, beamlet arrays, lines, arrays, geometries, and the like, as well as combinations thereof. The structured light may have a pre-specified position of diffraction orders such that the zero diffraction order has substantially the same intensity as the other diffraction orders.

With respect to fig. 5A-5B, the diffractive optical element 10 may be a composite optical material including a first optical material 12a and a second optical material 12B. The composite may have a first outer surface 16a opposite a second outer surface 16 b. The microstructures 11 may be formed along the first outer surface 16a of the first composite material 12a and then bonded to the inner planar surface of the second optical material 12 b. The second outer surface 16b of the second optical material 12b may be flat. In this way, the diffractive optical element 10 having the composite optical materials 12a, 12B (fig. 5A-5B) may have the same design as the diffractive optical element 10 having a single optical material 12 (fig. 4).

In one aspect, the second optical material 12b may be a material having a greater relative stiffness than the first optical material 12a, such as a rigid plastic or glass. The first optical material 12a including the microstructures 11 may be fixed to the second optical material 12B as shown in fig. 5B. In one aspect, an optical liquid adhesive (not shown) may be used to secure the first and second optical materials 12a, 12b to one another to form a composite optical material. Other means of attachment may be used so long as the composite of optical materials has microstructures 11 along the first outer surface 16a of the first optical material 12a and optionally no microstructures 11 along the second outer surface 16b of the second optical material 12 b.

In another aspect, the diffractive optical element 10 can include microstructures 11 along the first surface 14a of the optical material 12 and microstructures 17 along the second surface 14b of the optical material 12, as shown in fig. 6. Microstructures 11 and 17 may have the same or different phase profiles, such as a 2 π P unwrapped phase profile. In one aspect, the second surface 14b can have microstructures 17 that are identical (e.g., have the same phase profile) to the microstructures 11 along the first surface 14a of a single optical material 12. In another aspect, the second surface 14b may have microstructures 17 that are different (e.g., have different phase profiles) than the microstructures 11 along the first surface 14a of a single optical material 12. The microstructures 11, 17 may be formed as the optical material 12, as described above, or may be formed from two different pieces of the same optical material and then secured to each other in the manner described above.

In another aspect, the diffractive optical element 10 may be a composite of the first optical material 12a and the second optical material 12 b. The first optical material 12a may include a first outer surface 16a that includes microstructures 11. The second optical material 12b can include a second outer surface 16b that includes microstructures 17. The second outer surface 16b may have the same microstructures 17 as the microstructures 11 along the first outer surface 16a of the composite optical material. In another aspect, the second outer surface 16b can have a different microstructure 17 than the microstructure 11 along the first outer surface 16a of the composite optical material.

The diffractive optical element 10 in fig. 6 has two structured light patterns, a first structured light pattern from microstructures 11 along surface 14a of optical material 12 and a second structured light pattern from microstructures 17 along second surface 14b of optical material 12. The first structured light pattern may be the same as or different from the second structured light pattern. The projected light from the diffractive optical element 10 may be a combination of two structured light patterns with similar or dissimilar properties.

The phase profile of the microstructures 11 may be at least partially phase unwrapped to reduce phase discontinuities along the surface of the optical material 12. On the other hand, the phase profile of the microstructure 11 may be less than full (100%) phase unwrapping. The unwrapped phase profile may provide a zero diffraction order of the plurality of diffraction orders that may have substantially the same intensity as other diffraction orders of the plurality of diffraction orders.

The microstructures 11 along the surface 14a may have a phaseContour profile

Wherein phi_PIs the 2 π P-unwrapped phase profile, n is the refractive index of the surface, λ₀Is the center wavelength, and P is an integer selected to minimize the intensity of the zero diffraction order within the criterion of uniformity intensity error of the diffraction order of the structured light output from the device such that the light intensity in each of the plurality of diffraction orders is substantially the same. In one aspect, the phase profile is below full unwrapping in order to minimize uniformity or intensity errors at or within such criteria. The phase profile may be along one or two dimensions.

The phase profile defining the microstructure 11 may be periodic along the surface 14 a. The surface 14a extends in one dimension, but may extend along two orthogonal dimensions of the length and width of the surface 14a, both orthogonal to the depth dimension of the phase profile. In reducing the phase discontinuity until an acceptable uniformity of intensity error is achieved, while zero diffraction order hot spots can be avoided, the maximum depth dimension increases with each successive 2 π unwrapping of the phase profile.

The diffractive optical element 10 can be designed such that a specific value λ is assumed when the wavelength of illumination is assumed₀The diffractive optical element 10 can efficiently project structured light. For wavelength λ ≠ λ₀Because the phase relationship required for optimal image projection is no longer valid, performance can be severely degraded. As previously mentioned, the main feature observable in nearly all diffractive optical elements operating at wavelengths other than the design is the increase in the intensity of the zero diffraction order relative to the average intensity of the pattern of interest. In addition to operation at design wavelengths, as the angular spread, field of view (FOV), of the projected diffraction pattern increases, it becomes very challenging to ensure accurate fabrication of the theoretical surface profile. For example, small variations in the phase profile (such as phase depth errors) may make it almost impossible to minimize the zero diffraction order enough to make a diffractive optical element practical.

It has been shown that if the depth of the surface relief defining the diffractive optical element is increased, it is possible to operate the diffractive optical element at several specific wavelength values (typically 1 or 2 values). The rationale for this approach is based on the understanding that: as the total depth increases, the phase relationships necessary to ensure proper operation of the diffractive optical element can be satisfied, as in "Integrated differential and reactive elements for spread mapping" by s.noach, a.lewis, y.arieli and n.eisenberg, application.opt.35, 3635-. However, while the reasoning is correct, the increase in total depth is neither a necessary nor sufficient condition for acceptable multi-wavelength performance. Using this approach, it has been shown that images can be projected satisfactorily at two different wavelengths (i.m. barton, p.blair and m.taghizadeh, "Dual-wavelength operation differential phase elements for patterning," Opt, Express 1, 54-59(1997)), the disclosure of which is incorporated by reference in its entirety. In this case, the surface relief needs to use 16 phase levels, and the image is formed off-axis (off-axis) so that the zero diffraction order is not part of the projected structured light (image). One important reason for shifting the image is to avoid the use of the zero diffraction order, which tends to increase significantly in intensity when operating at several wavelengths. This may be due to the design itself not being optimized to suppress zero diffraction orders or manufacturing errors, which always reduces performance. However, this method cannot be used in applications such as 3D sensing.

There are several strategies available for generating diffractive optical elements that project arbitrary images. These strategies are suitable for monochrome operation and produce a phase structure with a total phase depth equal to 2 pi radians. In the case of a binary phase profile, the maximum phase depth is π, but the projected image is no longer arbitrary and needs to be centrosymmetric. As described above, the previous work enabled the design of elements suitable for projecting images under dual wavelength operation. For this reason, the maximum phase depth may be increased above 2 π radians. While this approach is convenient for dual wavelength operation, it has at least two serious limitations. First, it does not allow precise control of the intensity of the zero diffraction order. In this case, the usual solution is to shift the image to separate it from the zero diffraction order. As a result, if the intensity of the zero diffraction order increases, the main pattern is not affected, although it certainly reduces the efficiency, since a significant portion of the energy is concentrated on the zero diffraction order. However, this option is not available in cases where the presence of a strong zero diffraction order cannot be tolerated, such as for optical sensing. Secondly, depth diffractive elements often present serious difficulties for manufacturing, for example if the phase function is defined as a multi-order function. In this case, manufacturing may require exposing several masks, and accuracy should be guaranteed with respect to each mask, the relative depth of each elemental mask, and the alignment of the mask to the mask. Binary solutions are most common if the FOV is wide enough (typically greater than the full width of 20-30 degrees).

However, there is no method available for designing effective on-axis (on-axis) diffractive optical elements that operate at more than a single wavelength or over a wide FOV. The term "on-axis" refers to a diffractive optical element that can project an image that contains or is defined around a zero diffraction order. A simple example of such an element projection is an array of light spots, i.e. a distribution of image light spots in the far field arranged in an array. For a 3 x 3 array, an example of an array of spots is shown in fig. 2, which comprises zero diffraction orders as part of the pattern, as shown in fig. 3. In a more general case, the distribution of light spots is referred to as a "structured light" pattern and constitutes a specific distribution of light spots that can accommodate unique features, such as those required for detection in optical sensing applications.

For a more quantitative analysis, consider designing to project a total of N_sThe case of diffractive optical elements of individual spots. For simplicity, wide FOV or broadband operation has not been considered. One of the simplest ways to understand the occurrence of a strong zero diffraction order is by deviating the design phase depth. This may occur due to errors in manufacturing or operation deviating from the design wavelength. Both cases may be similar in the influence of the phase of the transmission. Thus, equivalence is used to examine problems operating at multiple wavelengths as a way of understanding sensitivity to depth without regard to any peculiaritiesThe manufacturing method is as follows. The phase depth error will be directly taken into account later. The structure of the element may be defined in terms of its phase delay, but for practical implementations, the actual physical structure needs to be defined. This is achieved by defining a design wavelength λ₀And a corresponding refractive index n (λ)₀) Is achieved by the refractive index n (lambda)₀) Associated with an optical material embodying a phase structure. If phi denotes the element at the design wavelength lambda₀The phase associated with operation at wavelength λ is given by α (λ) Φ, where:

where n (λ) is the refractive index at wavelength λ.

The term α may be referred to as a "wavelength detuning" coefficient because it measures the distance between the target design (α ═ 1) and actual operation. It also includes dispersion effects due to refractive index variations with wavelength. For purposes of illustration, consider an example of five operating wavelengths, such that α is 0.8,0.9,1,1.1, and 1.2 (or equivalently, these values may be considered as phase depth scaling errors). For example, if ignoring the band with λ₀The operating wavelengths will be 416.7nm, 454.5nm, 500nm, 555.6nm and 625nm, for a portion of the electromagnetic spectrum and dispersion of visible light of 500 nm. However, expressing the results as a is a more general approach, as it can refer to any specific region in the spectrum, not just the visible region.

Outside the ideal operating conditions, the performance of a diffractive optical element can be affected in several ways, however, the zero diffraction order is usually the most sensitive and easily perceived parameter. Other aspects such as reconstruction error and efficiency may also degrade, but they can generally be minimized, while for some image constraint sets, the zero diffraction order performance cannot be significantly improved. For these reasons, the selected performance measure is the ratio between the efficiency of the zero diffraction order and the average efficiency of the remaining spots in the structured light pattern, which is denoted by the greek letter ρ. If N is present in the desired pattern_sThe number of light spots is such that,the upper limit of the average efficiency per spot is 1/N_sIt is given. In practice, the actual diffraction efficiency may be below this value. Therefore, there is an upper limit for ρ. For simplicity, the projected image is considered to not utilize a zero diffraction order, meaning that the minimum energy is concentrated at the zero diffraction order in design. The diffractive optical element should have zero diffraction order efficiency at an ideally zero operating wavelength, which means that the values of ρ at α ═ 0.8,0.9,1,1.1 and 1.2 should ideally also be zero. In practice, these values will not be zero due to design itself or manufacturing errors.

Consider the case of a diffractive optical element with only two phase levels and a total phase depth of M pi, M being an odd natural number (1,3, 5.. eta.). This means that the element can only impart a phase delay of 0 or M pi radians. The ratio ρ of such elements is given by

The case M ═ 1 is a general binary diffractive optical element. It is also the easiest to manufacture because the total phase depth is only pi. The ρ values of the desired wavelength detuning (tuning) are shown in table 1.

Table 1: the ratio p at each wavelength detuning for a binary diffractive element projecting 100 spots according to the prior art. The two phase levels are 0 and M pi radians.

For most practical structured light applications, the total number of spots may be in the order of thousands or hundreds of thousands of spots. For example, a smaller number of spots may be used to illustrate the severity of the problem. Note that the best solution for the binary element, as shown in table 1, occurs at a total phase depth of pi, which seems to conflict with the concept that deeper phase delays lead to better multi-wavelength performance or equivalently low sensitivity to depth errors. The overall performance over a continuum between α -0.8 and 1.2 is shown in fig. 7. Once the equivalence of wavelength detuning and depth error is taken into account, the figure clearly illustrates the high sensitivity of the binary design, even with a small number of spots.

Moving beyond the binary solution, diffractive optical elements with multiple phase levels are considered. As a simple example, consider a 4-order diffractive optical element that again projects a structured light pattern of 100 spots, with zero diffraction order efficiency set to zero. In this case:

where the total phase depth is now 2 π M, and M is a natural number (1,2, 3....). Table 2 shows the performance of the 4-grade element. Compared to binary designs, 4-level designs exhibit a greater variety of behavior, but do not exhibit acceptable performance over all wavelengths or phase scaling errors of interest. As the elements become deeper, a narrow spectral range can be found where better performance is observed, but the absence of a single depth of structure enables acceptable performance for all specific values of a, let alone broadband operation. Again, these results may be further degraded by manufacturing errors, including mask misalignment.

Table 2: for an increased total phase depth 2 π M value, the element projects 100 spots (suppressing zero diffraction orders) at a ratio ρ of 4-order diffractive optical elements evaluated at several wavelength detuning values.

However, this approach provides hints as to the best design approach described later below: i.e. simultaneously increasing the number of phase levels and increasing the total phase depth. However, as seen above, simply increasing the total depth may not be sufficient to meet design performance criteria, and increasing the phase level may suffer from other difficulties, primarily related to manufacturing challenges, especially with the number of spots increasing in the wide FOV case. The invention provides a technical scheme which is far away from a diffraction optical element with discrete levels and takes continuously-changed gray-scale phase profiles into consideration. In this case, the constraint on the phase level can be eliminated, and the phase profile can take any value without limitation.

Before further movement, the following definitions are provided. The first step may be the construction of the structured light pattern. This refers to the number and distribution of spots of interest within a given FOV. Computationally, the projected image and phase function are defined as a matrix with R rows and C columns. The total number of pixels is then N — R × C. It is necessary to determine the complex transmission function of a diffractive optical element that provides a beam split propagating under the law of diffraction to generate the desired structured light pattern. For maximum efficiency, a phase-only solution is provided, although the procedure is generally also applicable to complex (phase and amplitude) transmission functions. The design process itself may be iterative, involving repeated propagation between the image and the beam splitter plane, with each step involving a correction factor, so that after a sufficient number of iterations a satisfactory solution can be found. Such iterative design methods are well known in prototype implementations described in Gerchberg and Saxton, pioneer works ("A practical algorithm for the determination of phase from image and diffusion plane pictures", R.W.Gerchberg and W.O.Saxton, Optik 35(2),237 and 246), the disclosures of which are incorporated herein by reference in their entirety.

An iterative design may provide a solution with a continuously varying phase delay for each pixel. In practice, even a gray scale phase profile involves some quantization, typically an 8-bit (256 phase levels) or 16-bit (65535 phase levels) pattern, which may be considered continuous for all practical purposes. By way of illustration, note that if Φ represents a phase function, then there is dispersionThe phase value is defined as phi _k2 pi k/(L-1), k 1,2, …, L, where L is the total number of levels (assuming now a total phase of 2 pi). Examples of phase quantization are shown in fig. 8A-8E. At 16 levels (fig. 8E), the quantization provides a good match to the original continuous contour. If the goal of phase quantization is to simulate a continuous profile as best as possible, then 8-bit depth (fig. 8D) or 16-bit depth (fig. 8E) may be effective.

As an interesting note, a three-level DOE is very effective in solving the zero diffraction order problem, as presented in "differential electronics design to supplied zero order product to surface depth error," ios standards in Optics and Photonics (TOPS), vol.75, differential Optics and Micro-Optics, Technical Digest, postcon Edition (Optical Society of america, washingdc, 2002), pages 58-60, and as disclosed in U.S. patent No.6,118,559, the disclosures of both of which are incorporated herein by reference in their entirety. In this case, the previously defined ratio ρ may be shown to be equal to:

wherein N is_eIs the number of phase levels multiplied by an even number pi (including zero), and N_oIs an odd multiple of the number of phase levels of pi. For example, if the phase function contains phase levels 0, π and 2 π, then N _e2 and N _o1. It is clear that only two combinations are possible: or N is_e＝N_oOr N is_e＝N_o+1. It is also assumed that the zero-order diffraction efficiency is set to zero by design. The graph in fig. 9 shows the performance in this case. Although 3-level solutions may in principle provide improved zero-order performance, the discrete nature of the pattern may still require multi-mask fabrication methods, which may be very challenging for large numbers of spots or wide FOVsIt is of war nature. Also, in the solution provided by the invention, the phase function of the diffractive optical element can be defined by a continuously varying phase profile, which can be manufactured without the need for a plurality of masks.

Thus, given a certain structured-light pattern, the phase function can be calculated via an iterative design method, according to the Gerchberg-Saxton method, or by other iterative designs as needed. These design methods exploit the continuous propagation between the phase plane and the image plane, mainly by using fast fourier transforms. The design period ultimately provides a phase function unit cell that is defined in the range-pi to pi for a total 2 pi phase shift. Since the absolute phase value is not of concern, only the total phase delay is considered. A technical solution that enables the diffractive optical element 10 of the present invention to be implemented without the problem of a zero diffraction order will now be described.

To increase the overall phase depth of the phase structure, a technique known as "phase unwrapping" is used. A typical iterative design method mathematically computes the phase function according to an arctangent operation. As a result, it can only provide phase values at intervals between-pi and pi at a given position. In general, the phase function at a given point (x, y) can be described by the general relationship:

Φ_P(x,y)＝Φ₀(x,y)+2πP(x,y), (6)

wherein phi₀Representing a phase diagram with phase values wrapped into a total 2 pi phase range. Function phi_PRepresenting a phase diagram with P (x, y) integers, where the phase is unwrapped by adding or subtracting integer multiples of 2 pi to remove as much discontinuity as possible in the phase function. For simplicity, P (x, y) is written as P, with the implicit understanding that it is a function of phase map position. The phase function derived from the iterative design method is represented by phi₀Given, and said that the calculated phase is wound modulo 2 pi. In other words, the total phase depth is limited to a total 2 π phase range.

Although unwrapping of the phase function has been used in optical design, it has not been applied to diffractive optical elements providing structured light until the present invention. For example, phase unwrapping has been used in diffractive tomography, as described in Devaney, U.S. patent No.4,562,540, in ophthalmic lenses with diffractive phase elements, as described in Apter et al, U.S. patent publication No.20100321635, and even in interferometry, as described in Bahk, U.S. patent No.9,921,111, for removing discontinuities in the phase function, the disclosures of all of which are incorporated herein by reference in their entirety.

For example, in "Optical performance of logical lenses," appl. opt.28, 976-983 (1989) of d.a. burally, g.m. morris, and j.r. rogers, the wrapping of the phase function to a modulo 2 pi or equivalent 2 pi wrap is described in defining a diffractive lens, the disclosures of which are incorporated herein by reference. This process is illustrated in fig. 10. In the case of a diffractive lens, the lens phase is deliberately wound to create a very thin element. The unique properties of a diffractive lens can be seen from this type of phase profile. In the case of a phase function of the structured-light pattern, the method typically yields a phase function of phi₀The phase of the 2 pi wrap is given and the concept of the present invention is to actually unwrap the phase profile to obtain a deeper phase pattern, as opposed to what is done in the prior art.

Although the winding and unwinding operations are quite simple in the case of a diffractive lens, the calculation methods applied to phase unwinding, such as described in the following books, can be used for the more general phase structure of the diffractive optical element 10 that produces the structured light pattern: dennis c.ghigia and Mark d.ritt, Two-Dimensional Phase unwrapting: Theory, Algorithms, and Software, Wiley-Interscience, first edition, 1998, the disclosures of which are incorporated herein by reference. However, other phase unwrapping methods may also be used. To illustrate the concept of the present invention, consider the case of a one-dimensional phase profile for simplicity. The principle is the same as a general two-dimensional phase profile, but easier to visualize. Most importantly, the behavior of the zero diffraction order is similar in both cases.

As an example, consider the phase profile shown in fig. 11, which generates a very simple distribution of 41 spots or sub-beams. The diffraction pattern resulting from the ideal phase profile of fig. 11 is shown in fig. 12. However, once some degree of error is introduced, the desired performance is lost. This is illustrated in fig. 13 for a phase depth error of 5%. As expected from the previous discussion, the main result of the deviation from the ideal phase is a rapid increase in the intensity of the zero diffraction order. Degradation of the uniformity of the various diffraction orders of interest is also noted. A typical measure of uniformity error, σ, is given by the expression:

wherein, I_maxAnd I_minRespectively the maximum and minimum intensity values in the stage of interest. For this particular example, the theoretical uniformity is 1.68%. In the case of a 5% phase depth error, the uniformity error increases to 13.45%, excluding the zero diffraction order.

The unwrapped phase profile is shown in fig. 14. If the diffraction pattern is calculated with an unwrapped phase profile that includes a 5% phase depth error, the results shown in FIG. 15 are obtained. It is immediately noticed that the diffraction pattern no longer shows strongly isolated zero orders. It has been found that even if the total phase depth is at a maximum and thus it is expected to eliminate the strong zero order completely, it is not desirable to completely unwrap the phase profile. The reason is that while the deeper phase profile solves the zero diffraction order problem, it also increases the uniformity error. In the example of fig. 15, σ is 63.29%, which degrades significantly in uniformity compared to the original 2 π wound phase profile. However, if the same phase profile is considered but 4 π wound, the performance is significantly improved, as shown in FIG. 16. Now, the uniformity error is given by 18.66%, which is slightly worse than the original 2 pi wound phase profile at the same level of phase depth error, but this is an acceptable criterion for uniformity error depending on the particular application. The corresponding 4 pi wrap compared to the original 2 pi wrap phase profile is shown in fig. 17. In a more general case, the optimum degree of unwrapping needs to be determined and is generally dependent on the structured light pattern as well as other parameters, such as the FOV and type of phase error that the manufacturing method may introduce. However, this example shows a method for producing the diffractive optical element 10, as shown in the block diagram shown in fig. 18.

The method for generating a diffractive optical element may comprise: calculating a phase profile that yields the structured light pattern (step 21); unwrapping the phase profile of the structured light pattern to obtain a 2 π P unwrapped phase profile, wherein P is selected to add or subtract integer multiples of 2 π so as to reduce discontinuities in the 2 π P unwrapped phase profile (step 22); generating a 2 pi P unwrapped phase profile (step 25); and a microstructure is fabricated along the surface of the optical material based on the generated 2 pi P unwrapped phase profile (step 26). The method may also include defining a structured light pattern (step 20). The method may further comprise evaluating the performance of the obtained 2 pi P unwrapped phase profile (step 23). The method may also include optimizing P (step 24). The method produces a diffractive optical element 10, which diffractive optical element 10 can exhibit a zero diffraction order having substantially the same intensity as the other diffraction orders of the plurality of diffraction orders.

The method may include defining a structured light pattern to be generated by the diffractive optical element 10 (step 20). In one aspect, the number of spots in the structured light pattern and their distribution, design wavelength and FOV can be defined. Non-limiting exemplary structured light patterns include spots, dots, beamlets, lines, geometries, arrays, or combinations thereof.

The method may include calculating a phase profile for diffracting light in a plurality of diffraction orders to provide a defined structured light pattern (step 21). For example, the calculation involves using the Gerchberg-Saxton algorithm or one of its variants. The result of step 21 may be a fully wrapped phase profile, as shown in the previous example of FIG. 11, where FIG. 11 is the calculated phase profile φ₀Which is wound by 2 pi for a one-dimensional structured light pattern of 41 spots or diffraction orders.

The method may comprise unwrapping the calculated phase profile to obtain an optimal level of unwrapping. Using the iterative unwrapping method as described above, one or several cycles of unwrapping of the phase profile occur (step 22) and each cycle is evaluated (step 23). In the example of fig. 11, for a set depth error of 5%, see the diffraction pattern of fig. 15 calculated for the phase profile of fig. 14, or the diffraction pattern of fig. 16 calculated for the phase profile of fig. 17. Other errors than 5% may be set as desired.

The method includes optimizing the P value to minimize the zero diffraction order while providing uniformity comparable to a 2P unwrapped phase profile (step 24). In other words, P is selected by adding or subtracting integer multiples of 2 π so as to reduce discontinuities in the 2 π P unwrapped phase profile until the intensity (or power) of the zero diffraction order is substantially the same as the other diffraction orders in the plurality of diffraction orders.

In the example of a one-dimensional structured light pattern of 41 spots or diffraction orders, the results of steps 22 and 23 are shown by the solid line of the phase profile of fig. 17, which represents an at least partially unwrapped phase profile, since less than complete unwrapping is desired to avoid excessive uniformity intensity errors. The original phase profile of fig. 11 for comparison is shown as a dashed line in fig. 17 in order to illustrate the reduced discontinuity in the resulting 2P unwrapped phase profile. The profile of FIG. 17 is less than fully unwrapped (i.e., the function φ is₀Full de-entangling) as this would result in too large a uniformity error as shown in fig. 15. The more limited unwrapping as shown in fig. 17 results in satisfactory performance with minimal degradation of uniformity errors compared to the ideal solution where phase depth errors are present. In practice, additional error sources can be used to evaluate performance, with minimal degradation of uniformity errors, compared to the ideal solution where phase depth errors are present. In practice, additional error sources may be used to evaluate performance relative to those to determine the optimum degree of unwrapping. For example, corners and sharp-edged corners may also play a role in addition to depth errors. The reason for choosing a specific example of this one-dimensional pattern is simple, as it can be easily visualized and interpreted. However, the same process can be directly extended to more complex structured light patterns or two-dimensional geometries, the only difference being the initial phase function at step 21 and the unwrapping method of steps 22 and 23 for two dimensions.

Once the optimum P is identified at step 24, the method may include generating a 2 pi P unwrapped phase profile s at step 25 by using the following equation:

wherein phi_PIs a 2 pi P unwrapped phase profile. The same method can be applied in 1D or 2D geometries in case of complex structured light patterns, regardless of the number of light spots. Also, the example of a linear dot pattern is for easy visualization and clear interpretation, but different 1D or 2D structured light patterns may be defined at step 20.

The method may include fabricating microstructures along the surface of the optical material 12 based on the generated 2 π P unwrapped phase profile. This may result in the microstructure 11 along the surface 14a having a phase profile of 2 π P unwrapped by steps 22-24. The 2 π P unwrapped phase profile can diffract the input illumination into structured light having a zero diffraction order with an intensity substantially the same as the intensities of the other diffraction orders. The 2 pi P unwrapped phase profile defining the microstructure may be periodic along the surface 14a such that the 2 pi P unwrapped phase profile is repeated in each unit cell of the diffractive optical element 10. An example of a unit cell of a 3 x 3 array of cells is shown in fig. 2, however, the number of cells along the surface 14a depends on the particular application of the diffractive optical element 10.

Fabrication of a microstructure along the surface of the optical material having a 2 π P unwrapped phase profile determined from equation (7) at step 25 can be performed by a variety of techniques. For example, the fabrication may be direct laser writing as described in U.S. patent 6,410,213, the disclosure of which is incorporated by reference, which utilizes a focused laser beam to expose a photoresist that has been coated on a substrate. As the laser beam scans the surface 14a of the optical material 12, the laser beam may be modulated so that a continuous simulated surface is obtained after development. Laser writing can create a phase structure in the photoresist, which can then be transferred to other materials by other techniques (e.g., microreplication or electroplating). In this way, a durable master mold can be created to produce a greater number of parts, such as hot embossing, injection molding, or replication. A suitable replica of the photoresist or pattern may also be transferred into the further optical material 12 by a reactive ion etching process or ion beam milling or similar methods.

As previously described, the second surface 14b and the second outer surface 16b can have microstructures 17, and the microstructures 17 can be the same as or different from the microstructures 11 of the first surface 14a and the first outer surface 16 a. If microstructures 17 are desired, steps 20-26 can be repeated to provide surface 17 with the desired structured light pattern using input illumination, the structured light pattern being provided from surface 12. In the case where microstructure 17 may be a mirror image of microstructure 11, only step 26 fabrication is performed.

One of the main functions of a diffractive optical element may be to distribute the illumination over a specified spatial area, wherein each individual beam is associated with a specific direction or similarly position and intensity in space. The beam profile from a given input beam can be done with high efficiency and without the presence of zero diffraction order hot spots.

Diffractive optical elements may be used in many applications, including but not limited to, facial recognition in-vehicle monitoring (e.g., automotive), lidar, and the like.

A method for producing a diffractive optical element, comprising: calculating a phase profile that produces the structured light pattern; unwrapping the calculated phase profile of the structured light pattern to obtain a 2 π P unwrapped phase profile; generating a 2 pi P unwrapped phase profile; based on the generated 2 π P unwrapped phase profile, a microstructure is fabricated along the surface of the optical material. 17. The method of claim 16, further comprising defining a structured light pattern. 18. The method of claim 16, further comprising evaluating performance of the obtained 2 π P unwrapped phase profile. 19. The method of claim 16, further comprising optimizing P. 20. The method of claim 16, wherein the diffractive optical element exhibits a zero diffraction order having substantially the same intensity as other diffraction orders of the plurality of diffraction orders.

From the foregoing description, it will be apparent that a diffractive optical element for diffracting an input illumination beam into a structured light pattern has been provided. Variations and modifications of the optical elements and methods according to the present invention described herein will no doubt become apparent to those skilled in the art. The foregoing description is, therefore, to be regarded as illustrative rather than restrictive.