POSITION ENCODER

Abstract

A device for determining the position of a light source, including an optical mask with a periodic pattern casting a shadow on an imager placed at a multiple of the Talbot distance or at a fraction of the Talbot distance.

Claims

1. A positioning device comprising: a light source, having a component with a given wavelength, a mask carrying a pattern having a given period, arranged to cast an image of the pattern on an imager placed at a distance from the mask a logic circuit connected to the imager for acquiring the image of the pattern cast on the imager, and arranged to determine a position of the light source based on the acquired image, wherein the distance is equal or a multiple or a fraction of the Talbot length arising from the wavelength of the light source and from the period of the pattern.

2. The positioning device of claim 1, wherein the mask and the imager are placed on parallel planes.

3. The positioning device of claim 1, wherein the mask and the imager are bidimensional.

4. The positioning device of claim 1, wherein the light source is made of reflected light.

5. The positioning device of claim 1, wherein the light source and the imager are placed on different planes and/or the mask is between the light source and the imager.

6. The positioning device of claim 1, wherein the mask and the imager are in a fixed spatial relationship.

7. The positioning device of claim 6, wherein the mask and the imager are on opposite parallel faces of a transparent plate or an optical filter plate.

8. The positioning device of claim 1, wherein said mask is placed at said distance equal or a multiple or a fraction of the Talbot length, wherein sub-periods of the period of the pattern carried by said mask are observed.

9. The positioning device of claim 1, wherein the mask has an absolute positioning code, the logic circuit being arranged to recognize the absolute positioning code and determine an absolute position of the light source based on the absolute positioning code.

10. The positioning device of claim 9, wherein the imager is smaller than the image of the pattern such that, for all possible positions of the light source, the imager records a sub-region of the image cast by the pattern and the absolute positioning code is such that the sub-region recorded by the imager contains sufficient information for determining the absolute position of the light source.

11. The positioning device of claim 9, wherein the pattern of the mask is a repeating bidimensional arrangement of optical features and the absolute positioning code consists in missing or altered features in the repeating arrangement.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The invention will be better understood with the aid of the description of an embodiment given by way of example and illustrated by the figures, in which:

[0019] FIGS. 1 and 2 illustrate geometrically an effect determining the response and the performances of the optical encoders of the invention.

[0020] FIG. 3 illustrates schematically a position encoder with an improved resolution thanks to the increase of the lever arm effect.

[0021] FIG. 4 illustrates a variant of the invention including a transparent plate or an optical filter plate or an air gap between the mask and the imager.

[0022] FIGS. 5 and 6 illustrate a part of a mask with a missing feature, and one of the corresponding Talbot images.

[0023] FIG. 7 illustrates a possible absolute spatial code based on the strategic omission of certain periodic features.

DETAILED DESCRIPTION OF POSSIBLE EMBODIMENTS OF THE INVENTION

[0024] FIG. 1 shows schematically a measurement system as disclosed in WO2012007561. The system determines the position, or the elevation angle α or the zenith angle D of the light source 35 to the detector 40. The light source 35 could be an LED, a laser source or any suitable light source. The following, description treats the source 35 as a monochromatic one. While sensible, this is not a necessary feature: the invention could operate with a white or broad-spectrum source by inserting a suitable spectral filter in the light path. The source 35 could also be a source of reflected light; in a variant, the light is reflected by an object whose position is to be determined.

[0025] The detector includes a mask 60 carrying a pattern with a period ‘a’, and casts a shadow on the imager 20. The mask 60 could be a grating with a repeating arrangement of opaque and transparent zones, a phase grating, an arrangement of micro-lenses or any arrangement of optical features that produce a recognizable shadow, that is a recognizable distribution of illumination, on the imager 20.

[0026] The imager 20, for example a CMOS image sensor, records the spatial distribution of illumination at the distance ‘d’ beyond the mask 60 from which the angle α can be computed. It must be understood that, although the FIG. 1 represents a simplified one-dimensional mask, the mask 60 and the imager 20 can in fact be bidimensional, and the detector 40 can locate the source in a two-dimensional representation, for example by computing the altitude and the azimuth angles of the source.

[0027] The resolution of the measuring system 40 is determined by the ability of the imager 20 to measure a shift in the shadow of the mask 60. As shown in FIG. 2, the relation is the same as in a pinhole camera: the shift of the shadow is geometrically related to the zenith angle D and the distance ‘d’ between the mask and the imager by x.sub.2=d.Math.sin θ.

[0028] If the mask 60 were strictly periodical, the light distribution seen by the detector 20 would repeat identically at x.sub.2=n.Math.a for each n integer. The mask can include an absolute code superimposed to the periodic pattern that allows the absolute determination of the mask position.

[0029] This geometrical description disregards the diffraction effects that blur the shadow pattern and reduce its visibility. The effect of diffraction is more severe for high ‘d’ and small ‘a’ therefore, the resolution of the known systems cannot be improved by a simple choice of a long throw distance d with a very fine grating a.

[0030] FIG. 3 shows how the present invention exploits the Talbot effect to overcome the above limitations. The distance between the imager 20 and the mask 60 is equal or a multiple or a fraction of the Talbot length Z.sub.T=2a.sup.2/λ arising from the wavelength A of the light source and the period a of the pattern. Thanks to the Talbot effect, the illumination pattern on the imager is an image of the mask, without the blurring effect of the diffraction. The inventors have realized that, unexpectedly, this advantage is retained in this application despite the high values of the incidence angle θ (up to and beyond 60°) and the fact that the patterns, having an absolute code superimposed, are not strictly periodic.

[0031] In an example, a position sensor with the structure of FIG. 3 had an 850 nm light source and a 100 μm periodic grating. These dimensions yield a Talbot length Z.sub.T=23.5 mm. Placing the pattern at the Talbot distance Z.sub.T allows the imaging of a pattern shadow without any diffraction effect. At 1 m distance, a 100 μm lateral displacement of the light source induces a displacement of the shadow signal of 100 nm in the standard configuration and 2.4 μm in the Talbot configuration, increasing the detection resolution by a factor of 24. Moreover, placing the pattern at twice the Talbot distance (2Z.sub.T) provides an increase by a factor close to 50 whilst maintaining a good image quality. This lever-arm effect has been measured for various multiples of the half-Talbot distance, the enhanced measurements matching with the theory.

[0032] The inventors have experimentally acquired sharp images of the pattern at distances d=0.5 Z.sub.T, d=Z.sub.T, d=1.5 Z.sub.T, d=2 Z.sub.T. The quality of the image was better than that obtained at non-Talbot arbitrary distances (d=0.9 mm), while increasing the lever arm effect and thus the resolution. In a further advantageous variant of the invention the distance d is set to a regular fraction of the Talbot distance. At these positions, the imager receives a scaled-down image of the grating. For example, at d=¼ Z.sub.T and at d=¾ Z.sub.T, the Talbot image has a period that is one half of the period a of the grating 60 or, rather, the spatial frequency of the illumination field is doubled. At d=⅚ Z.sub.T the spatial frequency is tripled, and so on. This can be advantageous because, when the image acquired by the imager 20 is analysed, the precision of the measurement is relative to the spatial frequency. Moreover, the scaled-down image could be captured, advantageously, by a smaller detector and/or by a detector with smaller pixels. One may say that placing the imager at a Talbot-fractional distance is akin to multiplying the resolution of the grating 60, without diffraction effects.

[0033] FIG. 4 illustrates a possible compact implementation of the invention in which the distance between the mask and the imager is set by a glass plate or an optical filter plate 25. One may, for example choose a near-infrared source at λ=940 nm, a mask with a period a=50 μm, and choose the thickness of the plate 25 such that the imager is at a distance of d=¾ Z.sub.T=3a.sup.2/(2λ)=3.99 mm.

[0034] Without choosing a distance that is related to the Talbot length, fine pitch gratings are not normally advantageous, because of the diffraction smearing of the image. In this embodiment of the invention, the large distance E and the mask's fine pitch give excellent resolution to this realization. Another advantage is that the space between the mask 60 and the image 20 is homogeneous with no changes of the refractive indexes, which the inventors have found to be advantageous. The logic circuit 50 acquires and analyses the illumination pattern present at the imager 20 and determines the position of the light source, using suitable interpolation algorithms.

[0035] The detector has been successfully tested on large incident angles (θ>60°) with different mask patterns, wavelengths, and distances.

[0036] The mask 60 preferably includes an absolute code superimposed to the periodic repetition with period a to enable the determination of an absolute position. The absolute code always implies some deviation from periodicity, for example a missing feature or an altered feature in the repeating arrangement of the mask. The inventors have found that, the position of the missing or altered fracture can be detected in Talbot image and at least in many sub-images. FIG. 5 shows a fraction of a mask 60 with a repeating arrangement of opaque (or transparent) features 61 and one missing feature at position 65, which could be taken as a bit of information in an absolute position code.

[0037] FIG. 6 is the corresponding half-Talbot image, at a distance d=a.sup.2/λ where the shadows 63 have the same pitch a of the original pattern but are shifted by half a period. The omitted feature 65 shows as four deformed and weaker shadows 68 in the image and can be readily detected.

[0038] FIG. 7 shows a possible implementation of an absolute positioning code in the mask 60. Since the imager 20 is smaller than the mask and the incidence angle is not zero, the imager 20 sees only a sub-region 22 of the whole pattern. The absolute code consists in an arrangement of missing features 65 that for all possible positions of the light source the sub-region recorded by the imager contains enough information for determining the absolute position of the light source. In this example—which is not the only possible—the absolute code consists in pairs of missing holes along the “x” and “y” directions, and each pair has a unique distance. The absolute position information is encoded in the distance between two missing features, and the spacing between missing features is chosen such that, for any possible position of the light source, the acquired image includes at least one—or more—pair of missing feature along the “x” direction and at least one—or more—pair of missing features along the “y” direction.