Transcript
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 11, Number 18 (2016) pp 9403-9408 © Research India Publications. http://www.ripublication.com
Development of an Optical Channel for a Motorcycle Helmet-Mounted Display System A. B. Artishchev Art Business LLC, Roshchinskaya 4, Office 503, Moscow-1154191, Russia.
Abstract Possibilities for creating a helmet-mounted collimated display system are explored building on an analysis of available technologies for the development of display systems. Optical calculation of components is performed as part of the study.
CHOICE OF THE DISPLAY GENERATOR This study also featured a comparative analysis of microdisplays and micro-projectors built upon various solutions. The main selection criteria included brightness and operating angles of emission, and energy consumption. The Microvision PicoP Display Engine (PDE) [1] laser scanning projector proved to be the most suitable projector in terms of brightness and power consumption. The device has the following structure. Narrow collimated ray beams from three laser sources (red, green, and blue) are superimposed with the use of a specter-dividing optical combiner. The superimposed ray beam (with a diameter of 1 millimeter) falls onto a micromechanical scanner that consists of two micro-mirrors. The brightness of the color components of the superimposed beam is modulated by video signal. The device then scans the combined ray in two dimensions thus building an image on the display similar to electronic ray scan in a television picture tube. The beam is focused on a typical distance to the display of 0.5–1 meter. With beam diameter at 1 millimeter it can be considered collimated. The system does not envisage focusing at a distance less than 60 millimeters. The micro-projector in question has the lowest power consumption. Ray deflection angles at the PDE exit are 49.7 horizontally and 25.4 vertically.
Keywords: display system, projection system, micro-display.
INTRODUCTION Active research is currently underway to develop augmented reality devices. Application of such solutions in various areas of human life can markedly simplify acquisition of relevant information in the course of dynamic activities. For example, a person driving a vehicle needs to respond fast to any change in the dynamic situation and can obtain necessary information with the use of augmented reality devices without being distracted by the instrument panel. This is especially relevant for motorcycles. Because of the small size and high maneuverability of their vehicles, motorcycle drivers often find themselves in near-accident situations, in which of crucial importance are the completeness and delivery speed of information about the situation, which the driver uses to make up their mind about further actions.
OBJECTIVE The objective of this study is to develop an optical system (OS) for a motorcycle helmet-mounted display system (HMDS) based upon the selected micro-projector. A series of requirements for the HMDS optical arrangement apply to this work. Fields of view within the eye span (angular dimensions of the virtual image) — 18 horizontally and 13 vertically, 3:2 aspect ratio; geometric distortion of the image shape — not more than 10%. Because the optical system under development works with the human eye, another important OS parameter should be accounted for, namely, exit pupil diameter, which for designs of systems of this kind should be at least 7 millimeters. Application of smaller diameters of light beams around the eye pupil will result in a recurrent disappearance of parts of the visible virtual image (or the entire image) from the viewer’s field of view because of the eye-ball movement, which is unacceptable for the transmission of safety-critical information. A large pupil diameter makes it possible to simplify or even rule out individual adjustment of the device that is otherwise required due to differences in users’ eyes and other anatomic features.
STRUCTURE OF THE OPTICAL SYSTEM Safety specifications that apply to motorcycle helmets require that there should be no elements between the visor and the human eye. Therefore, the known solutions based upon beam dividers and additional mirrors placed next to the viewer’s eyes are inappropriate. The main idea of the developed product is the use of the visor as a component of the optical system. If the visor is considered in the reverse direction of rays (from the eye to the display generator), then in the device under development beams are redirected from the inner surface of the visor to the above-the-forehead area bypassing the head. The accepted angular field of view within the eye span is 18º×13, which, for reflection from a typical motorcycle helmet visor (or a flat surface) produces unacceptably large dimensions of optical details in the above-the-forehead area. In order to reduce the cross dimensions of the optical system, its pupil must be located close to the lens unit. To this end, the visor should correlate the pupil of the eye with the intermediate pupil and the main optics of the system. This can be achieved with the use of ellipsoid surface, with the exit pupil of the system in one of its focuses and the intermediate pupil in the other focus (see Figure 1). This way the intermediate pupil can be placed in the above-the-forehead
9403
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 11, Number 18 (2016) pp 9403-9408 © Research India Publications. http://www.ripublication.com area (Figure 1, item 4). This system envisages an intermediate image (Figure 1, item 5), which is aberrated and placed between the intermediate pupil and the visor. Because the inner surface of the visor in use is the off-axis fragment of the ellipsoid, the intermediate pupil and the intermediate image have an inclination to the chief ray for the center of the field of view.
In other words, in order to meet the requirements with regard to the formation of viewing angles within the eye span and necessary image quality, the following is necessary: To emit ray beams with the necessary values of numerical aperture and angles of chief rays from the secondary source to achieve the design OS parameters — field of view, pupil diameter, and its position in the eye span; To form aberrated intermediate image, while ensuring the correction of aberrations following the reflection from the concave visor surface. PDE emission is collimated; therefore, the OS under development is telescopic. The Lagrange–Helmholtz invariant applies to such a system, with the correlation of angular fields at the entry and exit of the OS and beam diameter of Dпп sin пп Dout sin out (1) where Dпп is beam diameter at the PDE exit, Dout is exit pupil diameter; αnn is the PDE scanning angle, αout are the angular dimensions of the virtual image. Given the accepted HMDS vertical angular field of view of 13 and scanning angle at the PDE exit of 25.4, the correlation requires that the system have magnification proportionate to 1/2х in the vertical plane section [2]. In this case, it follows from correlation (1) that the diameter of the HMDS exit pupil within the eye span will be approximately 2 millimeters, which is not sufficient. To overcome the Lagrange–Helmholtz invariant and achieve the required value of the exit pupil diameter a device increasing the numerical aperture of the beams should be employed. To place the corresponding optical elements it is necessary to have the real minimally aberrated intermediate image. The diffusers and clouded glass that are normally used in such cases envisage high energy losses and uneven emission indicatrix. The most reasonable solution in terms of spatial and energy parameter is the exit-pupil expander (EPE) designed and patented by Microvision [3]. The EPE forms the requisite aperture angle and ensures even pupil filling. The pupil expander operates based upon the use of two identical arrays of micro-lenses [4], which are placed relative to each other with high accuracy at a distance of one focal distance of the micro-lens [5]. The size of each micro-lens is commensurable to the size of a pixel (30 microns). Taken together, they represent a double raster that ensures image resolution of at least 848×480.
Figure 1. Elliptical visor operation scheme. 1 — theoretical ellipsoid shape; 2 — visor fragment; 3 — exit pupil; 4 — entry pupil; 5 — intermediate image; 6 — system of axes to describe the ellipse fragment with twodimensional 4th degree polynomials.
It is necessary to form an image to place the EPE. A picoprojector can be used for this purpose, together with an additional lens to form the real image in its rear focal plane. Such a lens must have diffraction-limited image quality within 450–640 nm (based upon the PDE technical characteristics). It follows from Figure 2 that the lens to work with the PDE scanner will have to be a pinhole lens, with its pupil placed on the scanning mirror module inside the PDE body. Given the
The challenge of creating an HMDS optical system comes down to the correlation of the PDE optical characteristics with the parameters of inclined intermediate image, which are imposed by the ellipsoidal reflecting surface of the visor.
9404
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 11, Number 18 (2016) pp 9403-9408 © Research India Publications. http://www.ripublication.com existing dimensional limitations of PDE and its design peculiarities, the focal distance of the scanner lens is 26.8 millimeters. The exit pupil is located at infinity — a telecentric pattern of the chief rays is created in the image space. With beam diameter at the PDE exit of 1 millimeter, the numerical aperture of beams in the rear focal plane is equal to
Na
Dbeam 0, 017 2 F 'lens
ratio in the fields of view of the HMDS and the scanner (3:2 instead of 16:9), as well as various values of the OS focal distance in the vertical and horizontal planes. The equivalent vertical focal distance is 62 millimeters. Because the diameter of the exit pupil of the device is established at 12 millimeters, formula (4) can be used to calculate the numerical aperture of beams at the EPE exit
Na
(2)
Dexitpupil 2 F ' lenspart
0,11
(4)
In order to form the intermediate pupil, a decentralized collecting lens was used (see Figure 3). The off-axis part of the collecting lens refracts the chief rays coming from the secondary source, thus creating the intermediate pupil in the axis area of the focal plane of the collecting lens. A virtual image is created, inclined relatively to the chief ray of the central beam (Figure 3, item 2), which subsequently simplifies the coordination of the inclined intermediate image with the EPE plane that is perpendicular to the chief rays.
Figure 2. Scanner lens. 1 — Entry pupil of the lens; 2 — PDE body design limitations; 3 — focal plane of the lens. Maximal scanning angle corresponds to the maximum angle of inclination during scanning for image diagonal
Because beam deflection angles at the PDE exit are equal to 49.7º×25.4, with the focal distance of the scanner lens of 26.8 millimeters, the size of the image that it creates can be determined from the formula
tan( )
y' F 'lens
(3) Figure 3. Collecting lens.
and equals 25.0×12.18 millimeters.
1 — object; 2 — image of object 1; 3 — exit pupil of the collecting lens. The exit pupil of the collecting lens is at infinity
The EPE is the secondary source, i.e. an object for the subsequent part of the optical system, whose objective is to correlate the image formed by the EPE with the intermediate image formed by the ellipsoid visor in the reverse course, and formation of the intermediate pupil.
To create an inclined and properly aberrated image (Figure 4, item 5), a reproduction lens was used, located in the intermediate pupil area (Figure 4, p 1). Because of its location in the intermediate pupil area, its cross dimensions are minimized. The lens projects the inclined virtual image created by the collecting lens, into the intermediate image (Figure 4, item 3), which is transformed into the collimated image after being reflected from the ellipsoid surface of the visor. To coordinate the incline of intermediate images, it is necessary to additionally turn the reproduction lens around the pivot point of the intermediate pupil by 15.
To determine the numerical aperture at the EPE exit it is necessary to calculate the equivalent focal distance of the subsequent part of the system, which can be found according to formula (3) from the required fields of view within the eye span (equal to 18º×13°) and the size of the vertical EPE raster (12.02 millimeters). The equivalent focal distance of the collimating part in the vertical span can be calculated exclusively of distortion. The horizontal field of view of the scanner is not used in full because of differences in the aspect
9405
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 11, Number 18 (2016) pp 9403-9408 © Research India Publications. http://www.ripublication.com Image quality for the 12-millimeter pupil can be presented as a spot diagram (see Figure 7) for the array of spots of the field presented in Figure 6. Aberrations do not exceed 6 minutes for full pupil. However, real viewing is possible with eye pupil diameter of approximately 4 millimeters. For this value of eye pupil diameter, which can be located in any full pupil area of the OS, aberration values are reduced four to eight times compared to the full pupil, i.e. approximately 0.75–1.5 angular minutes depending on the spot of the field and position of the eye pupil inside the pupil of the system. These values correlate with the eye resolution.
Figure 4. Reproduction lens. 1 — lens; 2 — plane mirror; 3 — object; 4 — virtual object (image of object 3 in mirror 2); 5 — image of object 3 through lens 1; 6 — principal planes of the lens. Figure 6. Positions of field spots in the focal plane of the scanner lens.
The optical system after aberration correction is presented in Figure 5.
Spots are symmetrical relative to OY axis.
Figure 5. Principle HMDS optical layout. 1— exit pupil of the optical system; 2 — fragment of the ellipsoid visor; 3 — reproduction optical system; 4— plane mirror refracting the optical axis; 5 — collecting lens; 6 — pupil expander; 7 — scanner lens; 8 — PDE; 9 — intermediate pupil; 10 — intermediate image.
Figure 7. Spot diagram of the quality of the HMDS optical system.
9406
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 11, Number 18 (2016) pp 9403-9408 © Research India Publications. http://www.ripublication.com
The 100 micron scale of aberration graphs in linear terms corresponds to 10 angular minutes in the eye span for a 12millimeter pupil. The HMDS OS is characterized by asymmetrical trapezoidal distortion, demonstrated in Figure 8. To ensure effective adjustment, relevant predistortion should be introduced to the shape of the lens. This can be done with the use of software and hardware correction of the PDE scanner, because the scanner field is not used in full.
Currently pico-projectors are available with matrix formats of 0.3 inches and 0.47 inches, with respective work areas of 6.9×3.9 millimeters and 10.4×5.8 millimeters. Because the aspect ratio of fields of view (3:2) differs from the aspect ratio of the work areas of the matrix (16:9), the work areas of the matrices will not be used along the bigger side. Areas in use will amount to 5.4×3.9 millimeters for the 0.3-inch matrix and 8.1×5.8 millimeters for the 0.47-inch matrix. Optical parameters of the projector part of the entire helmetmounted system can be calculated using formulas (3) and (4). Focal distances of the collimating system exclusive of distortion will amount to 17 millimeters for the 0.3-inch matrix and 25.5 millimeters for the 0.47-inch matrix. For a 12millimeter pupil the numerical aperture within the matrix span will amount to 0.35 and 0.23 for the 0.3-inch matrix and 0.47inch matrix, respectively. DLP (Digital Light Processing) projectors are subject to limitations on the aperture of beams. The maximum working numerical aperture (NA) of the matrix is equal to 0.29. Therefore, for a 0.3-inch matrix, the required angular field values cannot be achieved with OS pupil of 12 millimeters. The use of the 0.47-inch matrix has potential to increase angular fields of view to 26.6º×15.7° for the 12-millimeter OS pupil. However, when the field of view is increased to the said dimensions, the optical elements should be increased correspondingly, which may result in a situation, when the system cannot be placed inside a motorcycle helmet. All of the calculations mentioned above are exclusive of distortion and therefore approximate. The calculated values of the numerical aperture of beams within the matrix span suggest that decentralized and nonspherical optical elements should be used to correct asymmetrical aberrations. The objective of HMDS OS design therefore comes down to the search for compromise between the largest possible diameter of the exit pupil and field of view with high image quality, on the one hand, and dimensions and sophistication of optics, on the other hand.
Figure 8. Image distortion in the eye span exclusively of predistortions.
One of the advantages of the developed OS is that due to the sufficient margin of dimensions it was possible to minimize the number of decentralized elements. Furthermore, there are no non-spherical components, except for the ellipsoid surface of the visor. Its drawbacks include the large number of additional optical components, such as the scanner lens and EPE, which are necessitated by the use of the image generator of the selected scanning type. The length and dimensions of the optical arrangement can be reduced if another image generator is used.
CONCLUSION An optical system that fully meets the specified requirements has been developed within the framework of the study. Furthermore, current micro-projector technologies and developments were analyzed for use as a source of emission and output information. The system developed as part of the study has been implemented as a simulator, tested and showed its operability.
MEMS-based projectors can serve as alternative image generators, for example Texas Instruments pico-projectors, characterized by high brightness and contrast. In this case, primary image is formed on the matrix surface. In the optical layout, the matrix is placed in lieu of the intermediate image (Figure 5, item 6); therefore, there is no need for the scanner lens (Figure 5, item 7) or EPE pupil expander.
ACKNOWLEDGEMENT The study was performed with financial support from the Ministry of Education and Science of Russia; project identifier RFMEFI57614X0082.
9407
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 11, Number 18 (2016) pp 9403-9408 © Research India Publications. http://www.ripublication.com REFERENCES [1]
[2]
[3]
[4]
[5]
Freeman M., Champion M., Madhavan S. Scanned laser pico-projectors: seeing the big picture (with a small device). Optics & Photonics News. 2009. Vol.20, Number 5, pp. 28–34. Gan M., Starkov A., Larionov S. About a violation of the Lagrange-Helmholtz conditions in the optical display system. Proceedings of Applied Optics Conference 2012, pp.23-24. Powell K.D., Ürey H., Malik A., Hannigan R. J. Microvision. Scanned-beam heads-up display and related systems and methods. Patent US 7460305. USA. 2008. Urey H. Diffractive exit-pupil expander for display applications. Applied Optics. 2001. Vol. 40. № 32. pp. 5840–5851. Urey H., Powll K. Microlens-array-based exit-pupil expander for full-color displays. Applied Optics. 2005. Vol. 44. Number 23. pp. 4930–4936.
9408
