1. INTRODUCTION

It has been more than 70 years since the invention of holography [1]. This technology has fascinated many people and has been actively researched and applied in many fields. The main functions of this technology are wavefront recording and reconstruction, and control; wavelength dispersion; and wavelength and angular selectivity. These functions have been used in many applications such as 3D displays, interferometry, holographic optical elements, and holographic memory. Among these, 3D displays is where the holography is best used because true 3D images can be obtained through holograms [2].

Holograms are classified as transmission holograms or reflection holograms (RHs) on the basis of the direction of the reference beam and object beam [3]. In principle, a transmission hologram is illuminated from behind, and the image is observed from the transmitted beam. Although the resolving power of the recording material can be relatively low, a laser beam is basically required for image reconstruction. When illuminating the hologram with white light, it is not possible to reconstruct a sharp image with depth due to wavelength dispersion. On the other hand, a RH requires high-resolution recording material, but its excellent wavelength selectivity enables sharp images to be reconstructed with white light. Basically, the illuminating beam is applied from the front of the hologram, and the image is observed by the reflection from the hologram.

In principle, a 3D color image is obtained by recording a hologram with laser beams consisting of the three primary colors of R (red), G (green), and B (blue) [4]. The image gives a strong impression, as if the original object were there. To achieve this with a transmission hologram, it is necessary to devise methods of eliminating an unwanted image, i.e., a cross-talk image. Several methods have been proposed for this purpose, but they result in low signal-to-noise ratio, making it difficult to reconstruct a clear color image. An RH, however, has excellent wavelength selectivity, so there is no cross-talk image, and a brightly colored image can be easily reconstructed with white light [5,6]. High-resolution recording materials such as silver halide emulsions [7,8] and photopolymers [9,10] have been developed to reconstruct a bright image. Therefore, a RH has high utility value for 3D displays.

When comparing the two types of holograms from the viewpoint of a method for illuminating holograms, a RH requires a large space between it and the illuminating light source, and when the observer approaches the hologram to touch the image, it becomes a shadow of the observer and the image is lost. In a bright room, a hologram is illuminated with light other than the original illuminating light source, resulting in a ghost image. As a configuration of illuminating an RH from behind, using a beam splitter (BS) has been proposed [11]. The illuminating beam passes once through the RH and hits the BS, and the hologram is illuminated with the reflected beam from the BS. This configuration is simple, but the transmitted light leaks to the observer’s side, and as in the case of illumination from the front, the hologram is illuminated with the light in the space on the observer’s side. Therefore, an unwanted image is superimposed, and reflected light from the BS is generated.

We propose a system that utilizes an RH to realize color holography that illuminates from the back site. This configuration combines polarizers and quarter-wave plates (WPs), which can mitigate the drawbacks mentioned above and enable reconstruction of clear color images. We discuss the incident angle dependence of the polarization characteristics of the elements used, a benchtop prototype of the system we constructed, and the obtained reconstructed image.

2. CONFIGURATION OF THE SYSTEM

Figure 1 shows the conceptual diagram of the proposed configuration. The configuration consists of two absorption-type polarizers (APs), one reflective polarizer (RP), and two WPs. An AP is a device that allows any linearly polarized beam to pass through. It can be a general polarizing film, such as that used in LCDs. An RP is a device that passes through an arbitrary linearly polarized beam and reflects an orthogonal linearly polarized beam. Since we have color holograms in mind, these elements are broadband elements covering the entire visible region. The white light illuminating beam incident from the right side at a certain angle is first circularly polarized using AP1 and WP1 and passes through the RH. The beam is then reflected using the RP and passes through WP2 again to illuminate the RH. The reconstructed beam from the RH passes through WP2, RP, and AP2 to the observer.

 

Fig. 1. Configuration of the proposed system. AP1, AP2, absorption-type polarizer; WP1, WP2, quarter-wave plate; RH, reflection hologram; RP, reflective polarizer; O, observer.

Download Full Size | PDF

AP1 and AP2 are the same element, and WP1 and WP2 are the same as well. The illuminating beam is a collimated beam, and its angle is the same as that of the reference beam used in recording the RH, which is usually in the range of 30° to 60°. AP1 makes the illuminating beam linearly polarized (parallel to the incident plane, p-polarized), and the two WPs convert its polarization direction by 90° to s-polarized. The RP efficiently reflects the s-polarized beam and acts as a reflector for illuminating the RH, and at the same time, has the effect of blocking the transmitted beam and preventing the illuminating beam from leaking to the observer side. WP2 passes twice between the reflections of the RP and RH to further convert the polarization direction by 90° to become a p-polarized beam, and efficiently passes the subsequent two polarizers.

On the other hand, the light reconstructed from external light in the space on the observer’s side, such as the illumination light of the ceiling, is blocked by the RP because the polarization direction of the light passing through the RP is converted by 90° by using WP2, thus eliminating the effect of the reconstructed light caused by external light. AP2 prevents the s-polarized light component of the external light from being reflected by the RP.

The combination of APs, WPs, and RP makes it possible to illuminate the RH from the back site and to observe a clear reconstructed image without the disruptive transmitted light or reflected light caused by external light, which is the case with using a BS.

3. DEPENDENCE OF INCIDENT ANGLE ON POLARIZATION CHARACTERISTICS OF ELEMENTS

An RH is usually reconstructed by illuminating it at the same angle as the reference beam. Therefore, the illuminating beam is obliquely incident on the elements. Since the object beam spreads out vertically and horizontally, the reconstructed beam also passes through the elements diagonally. Therefore, we measured the polarization characteristics of each element for an obliquely incident beam. Regarding the wavelength of the reconstructed beam from the RH, only a narrow width centered on the wavelength when it was recorded is reconstructed. Therefore, R, G, and B laser beams, which are the same as the recording wavelength of the hologram, were used for this measurement. These wavelengths were R, 639 nm; G, 532 nm; and B, 459 nm. The elements used are as follows.

Figure 2 shows the optical layout for measuring the polarization characteristics of AP1. The linearly polarized collimated beam from the laser is used as an incident beam. The transmitted beam from AP1 is measured for R, G, and B when it is tilted at an incident angle $\theta x$ to the $z$ axis in the $x {-} z$ plane. For convenience of measurement, the incident beam was fixed, and AP1 was rotated. Similarly, the characteristics were measured for tilted angle $\theta y$ with respect to the $z$ axis in the $y {-} z$ plane.

 

Fig. 2. Optical layout for measuring polarization characteristics of AP1. D, detector; $I{{0}}$, intensity of incident beam; $I{{1}}$, intensity of transmitted beam.

Download Full Size | PDF

Figure 3 shows the dependence of the transmittance (intensity ratio of the transmitted beam to the incident beam) on the incident angle $\theta x$. The transmission axis of AP1 is in the $\pm 60 ^\circ$-axis direction, and APxT is the case in which the polarization direction of the incident beam is the same and is p-polarized. The polarization of the beam with the same polarization direction as the incident beam is maintained for oblique incidence up to about $\pm 60 ^\circ$ for R, G, and B, and the transmittance is more than 75%. On the other hand, APxS is the case in which the polarization direction of the incident beam is s-polarized. The amount of the transmitted beam in this case is ideally zero, but it does exist. However, the amount is less than 1%, and the leaked beam in this direction is almost cut off. This indicates that this configuration is effective for a reference beam up to an angle of about $\pm 60 ^\circ$, and for a certain amount of divergent beam, not collimated beam. The angle of the object beam is also effective up to this extent.

 

Fig. 3. Dependence of the incident angle on polarization characteristics for AP1 (SEG1223CUHC film). The angle of the incident beam changes in the $xz$ plane. $\theta y = {{0}}$. APxT, transmission axis is the same as that of the polarization direction of the incident beam; APxS, transmission axis is perpendicular to that of the polarization direction of the incident beam. The plots of APxS-R and APxS-G are hidden behind APxS-B.

Download Full Size | PDF

Similar measurements were carried out for $\theta y$. The results are shown in Fig. 4. This is because the reconstructed beam spreads not only in the $x$ direction but also in the $y$ direction, so it is necessary to investigate the polarization characteristics of the beam incident obliquely in the $y$ direction. APyT is the case in which the transmission axis of AP1 and the polarization direction of the incident beam are the same. As the angle of incidence increases, the transmitted beam in this polarization direction simply decreases. The transmitted beam at $\pm 60 ^\circ$ is about 70% or more of that of the vertical incident. APyS is the case in which the polarization direction of the incident beam is orthogonal. The transmitted beam in this case is almost completely cut off in all R, G, and B. From the above results, it was found that the polarization performance of the APs used with respect to the incident angle was maintained up to about $\pm 60 ^\circ$ in both the $x$ and $y$ directions.

 

Fig. 4. Dependence of the incident angle on the polarization characteristics for AP1 (SEG1223CUHC film). The angle of the incident beam changes in the $y {-} z$ plane. $\theta x = {{0}}^\circ$. APyT, transmission axis is the same as that of the polarization direction of the incident beam; APyS, transmission axis is perpendicular to that of the polarization direction of the incident beam. The plots of APyS-R and APyS-G are hidden behind APyS-B.

Download Full Size | PDF

The polarization characteristics of the WPs were measured in the same manner by stacking WP1 and WP2 with the optic axis in the same direction instead of AP1, as in Fig. 2. The polarization direction of the incident beam was the $x$-axis direction, and the direction of the optic axis of the WPs was set at an angle of 45° from the $x$ axis. The purpose of using two WPs is to make the polarization direction of the transmitted beam orthogonal to the incident beam. To confirm this, A polarizer, which is the same type as AP1, was inserted between WP2 and the detector (D), and the transmitted beam was measured.

Figure 5 shows the results of measuring the dependence of $\theta x$ on the transmittance. WxT is the case in which the polarization direction of the AP is set to be orthogonal to that of the incident beam. As can be seen from the figure, B is slightly inferior to R and G, but in each case, the transmittance is about 50% or more up to an incident angle of $\pm 60 ^\circ$, and it remains more than 75% with respect to the case of vertical incidence. Beyond this angle, the transmittance decreases rapidly. WxS is the case in which the polarization direction of the AP is the same as that of the incident beam, and it can be used as a guide to how much the polarization component of the incident beam exists. At vertical incidence, the transmittance is less than 1% for R, G, and B, and gradually increases as the angle increases. At 60°, it is 4%–7%. The dependence of the transmittance on $\theta y$ is almost the same as that of $\theta x$. The reason for inserting the polarizer was to extract the polarization component in a specific direction. Its transmittance in the direction of the transmission axis is 75% to 80% for vertical incidence, as shown in Fig. 3. Therefore, considering the attenuation at the polarizer, the transmitted beam after passing through WP2 is about 30% larger than the measured value. From these results, the effect of the WP is maintained up to an incident angle of about $\pm 60 ^\circ$ in both the $x$ and $y$ directions.

 

Fig. 5. Dependence of the incident angle on the polarization characteristics for WP1 (NZF film). The angle of the incident beam changes in the $x {-} z$ plane. $\theta y = {{0}}^\circ$. WxT, polarization direction of an AP is perpendicular to that of the incident beam; WxS, polarization direction of the AP is the same as that of the incident beam.

Download Full Size | PDF

Similar measurements were carried out for the RP. As shown in Fig. 1, the beam incident on the RP is s-polarized. One of the functions of the RP is to efficiently reflect the obliquely incident beam with this polarization direction and direct it to the RH. To investigate the effect of the RP, the $\theta x$ dependence of the intensity $I$ 1 of the reflected beam from there was measured by changing the polarization of the incident beam to s-polarization and replacing AP1 with RP. The transmitted beam leaking from the RP was also measured. The results are shown in Fig. 6. The reflectance ($I{\rm{1/}}I{{0}}$) for vertical incidence was taken as the average of the reflectance at $\pm 3 ^\circ$. RPxR is the case in which the polarization direction of the incident beam to the RH and the reflection axis of the RP are the same, and RPxS is the case in which they are orthogonal to each other. The reflectance is more than 80% up to the incident angle of $\pm 60 ^\circ$, and the leaked beam was less than 0.3% for R, G, and B. Therefore, the RP also showed sufficient performance.

 

Fig. 6. Dependence of the incident angle on the polarization characteristics for RP (WGFfilm). The angle of the incident beam changes in the $x {-} z$ plane. $\theta y = {{0}}^\circ$. RPxR, reflection axis is the same as that of the polarization direction of the incident beam (s-polarized light); RPxS, reflection axis is perpendicular to that of the polarization direction of the incident beam. The plots of RPxS-R and RPxS-G are hidden behind RPxS-B.

Download Full Size | PDF

Another function of the RP is to efficiently pass the p-polarized light reflected by the RH and through WP2. To investigate this effect, measurement was carried out in the same arrangement as when the characteristics of AP1 were measured. The results indicate that when the transmission axis is the same as that of the polarization direction of the incident beam, the transmittance of R, G, and B is maintained above 70% up to about $\pm 60 ^\circ$, and the leaked beam is less than 0.5%. For the $\theta y$ dependence, the transmittance decreases as the angle of incidence increases, as in the case of AP1 but does not decay as much and remains above 70% at $\pm 60 ^\circ$.

The total transmittance of the proposed configuration was estimated from the results of these measurements for the APs, WPs, and RP. Assuming that the transmittance and reflectance of the RH are both 100%, the total transmittance is calculated to be about 40% for R and G and about 30% for B for vertical incidence. As the angle of incidence increases, the transmittance decreases, but in the case of incidence of $\pm 60 ^\circ$ in both $x$ and $y$ directions, the transmittance decreases only about half that of vertical incidence.

From the above results, it was found that this configuration has sufficient polarization performance for an incident beam up to about 60° in both directions and can reconstruct an image with a wide view range. Recent digital hologram printers have made it possible to record a color image hologram with a full parallax of 120° [12]. The proposed configuration can also be used for illuminating such holograms.

4. EFFECT OF BACK-LIGHT ILLUMINATION ON THE RH RECONSTRUCTED BEAM

In the proposed system, the beam is incident from the back site of the RH. The direction is symmetrical to the reference beam used for recording with respect to the surface of the hologram. This may result in the reconstruction of a true or a conjugate beam that corresponds to a true image or a conjugate image that is different from the originally desired reconstructed beam. When these unwanted beams are reconstructed, a portion of the illuminating beam is used, which reduces the amount of light transmitted through the RH. The transmitted beam then becomes the illuminating beam of the RH after folding back by the RP. Ideally, the beam should not be reconstructed by the illumination from the back site, which is different from the conventional illumination system, but this may occur depending on the angle of the object beam. To confirm this, we analyzed whether the unwanted beams are generated and whether the beams affect the transmitted beam. The former depends on the angle of the reference and object beams, and the latter is also related to the angular and wavelength selectivity of the RH. To analyze these things, we used a simulator [13] for analyzing the imaging characteristics [14], including the diffraction efficiency [15] of a hologram.

In the following, the unwanted beams that may affect the transmitted beam are discussed for the true and the conjugate beams.

 

Fig. 7. Optical arrangement for analyzing behavior of the beam reconstructed with the RH when it is illuminated from back site. (a) Recording, (b) reconstruction.

Download Full Size | PDF

The optical arrangement for the analysis is shown in Fig. 7. The RH is recorded using laser light with three wavelengths: R (639 nm), G (532 nm), and B (459 nm). A photopolymer with a thickness of 16 µm is used as the recording material. In consideration of the proposed system, the angle parameters are set as follows:

We first discuss the effect of the true beam on the transmitted beam. In Fig. 7, the illuminating beam of the RH is incident at $\theta cx = {{45}}^\circ$ from the back site. This angle is the same as for the object beam with $\theta ox = {{45}}^\circ$ and $\theta oy = {{0}}^\circ$. In this case, the direction of the true beam reconstructed at the same wavelength as the recording is the same direction as the reference beam of $\theta rx = {{45}}^\circ$ for each component hologram, and the Bragg condition is satisfied for this beam, so the unwanted beam is reconstructed in this direction. Then, if the diffraction efficiency of the beam is high, the intensity of the transmitted beam decreases. However, the object beam exists in all directions, and if the diffraction efficiency of the reconstructed beam is the same for all directions, the contribution of a very narrow direction around 45° is limited and slight because of excellent wavelength selectivity of the hologram. The full width at half-maximum of the wavelength of the reconstructed beam is about 6 nm, and the corresponding angular width in the $x$ direction is about 5° for a G-component hologram. This angular width is approximately the same when the $y$ component is added to the direction of the object beam. Estimating from these, the attenuation of the transmitted beam is about 4% considering the total angular width of the object beam. Assuming that the overall diffraction efficiency of RH is 80%, the amount is only about 3%. This value varies slightly, depending on the recording wavelength, and is about 4% even for the most affected R-component hologram. Therefore, the effect of the loss of the transmitted beam due to this matter is almost no problem.

For holograms reconstructed at wavelengths different from those of the recording (for example, 639 nm and 459 nm for G-component holograms), the intensity of the reconstructed beam is very weak because the Bragg condition is not satisfied for any incident angle of the object beam.

We examined whether there are any other unwanted beams in the true beam, because the illuminating beam is white light, and the object beam is assumed to exist in all directions. The unwanted beams that satisfy the Bragg condition exist even at angles other than the incident angle of 45°. However, the wavelength of the beam is different from the three wavelengths at the recording. The light of these wavelengths affects the attenuation of the transmitted light but does not affect the light of the three wavelengths R, G, and B required to reconstruct the RH.

For the conjugate beam, it is far from the Bragg condition for the incident object beam in all directions, and the intensity of the reconstructed beam is extremely low.

From the result of the above consideration, the influence of the back illumination on the transmitted beam does not need to be considered practically in the range where the incident angle of the object beam is within $\pm 60 ^\circ$ for all directions.

5. PROTOTYPE OF THE PROPOSED SYSTEM

We constructed a benchtop prototype of the proposed system. Since each element is a thin film, the configuration has two components. Component 1 is composed of the SEG1223CUHC film (AP1) and NZF film (WP1) laminated onto a 2-mm-thick transparent glass plate in this order from the incident beam side. The optic axis of WP1 was set at 45° to the transmission axis of AP1. Component 2 is composed of the NZF film (WP2), WGF film (RP), and SEG1223CUHC film (AP2) laminated on a 2-mm-thick transparent glass plate in this order from the incident beam side. The RP and AP2 were laminated so that their transmission axes were aligned and the optic axis of WP2 was at 45° to the RP axis.

Component 1, the RH, and component 2 are arranged in this order from the incident beam side. The RH is covered with a glass plate for protection, and the total thickness, including the glass plate, is about 5 mm. Therefore, the thickness of the configuration including all elements is approximately 10 mm. This thickness will remain the same even if the size of the hologram increases, as long as the thickness of the support does not change.

To confirm the effectiveness of the prototype, it was illuminated with the original illuminating light source and an external light, and the quality of the reconstructed image were observed. A light-emitting diode device (HOLOLIGHT) manufactured by PiPHOTONICS Co., Ltd. [16] was used as the original illuminating light source. Figure 8 shows a photograph of the prototype including HOLOLIGHT located at the bottom. From HOLOLIGHT, almost all the collimated light comes out upwards. The system in the flame is placed vertically and illuminated at 45° from above through a mirror set on top. The total height of the benchtop prototype of the proposed system is 82 cm.

 

Fig. 8. Benchtop prototype of proposed system.

Download Full Size | PDF

Figure 9 shows an example of a reconstructed image obtained from this prototype. One of the advantages of this system is that it is not easily affected by external light. Figure 10 shows the effect on the reconstructed image when there is a fluorescent lamp mounted on the ceiling as well as HOLOLIGHT. Figure 10(a) shows the image when the hologram was directly reconstructed by the conventional system. The reconstructed image reconstructed from the fluorescent lamp overlaps the original reconstructed image as a ghost image, degrading image quality. Figure 10(b) is the image reconstructed with this proposed system under the same lighting conditions. There is no ghost image, resulting in a clear image.

 

Fig. 9. Example of reconstructed image of color hologram using proposed system.

Download Full Size | PDF

 

Fig. 10. Effect of external light on the reconstructed image when fluorescent lamp was mounted on ceiling. (a) Reconstructed image with conventional system; (b) reconstructed image with proposed system.

Download Full Size | PDF

Each element used to fabricate the prototype does not exhibit the ideal perfect properties, as experimentally confirmed. However, by using each of the adopted elements, it was possible to confirm the sufficient effect of the proposed configuration, as shown in Figs. 9 and 10.

Regarding the lighting source, by combining a small light source and a short focus lens or a concave mirror and arranging them in an array, a thin and compact illumination device can be realized, and a large hologram is possible to illuminate.

6. CONCLUSION

We proposed a system that utilizes an RH to realize color holography that illuminates from the back site. Compared to the conventional system of illuminating from the front, our system has the following advantages: it does not require a large space between the illuminating light source and hologram and does not generate ghost images caused by light other than the illuminating light source. Regarding the dependence of the polarization characteristics on the angle of incidence of the APs, WPs, and RP, which are the elements constituting this system, it was found that the polarization performance was maintained for an incident beam of up to about $\pm 60 ^\circ$ in the vertical and horizontal directions, making it possible to observe clear color images without ghost images over a wide viewing range. The hologram illuminated with this system is a conventional RH and does not require any special modification, so its utility value is high. This system is expected to be used for space-floating 3D displays and other applications that take advantage of its features.