1. Introduction

In general, recording of digital holograms is achieved optically using an array photo-sensor [1], while the captured image can be reconstructed numerically or optically. Numerical reconstruction opened new possibilities in solving problems such as the suppression of the zero-order disturbance [2–4], compensation of the aberrations [5,6], dynamic measurements [7–10], detection of hidden deformations in vibration patterns [11,12], or multiplexing and demultiplexing of information for remote reconstruction [13]. Recently demonstrated optical reconstruction of digital holograms with coherent light introduced additional features [14] implemented also into the real-time systems [15]. However, optical reconstruction with coherent light suffers from the known speckle noise especially when realized with the liquid crystal displays with large pixels [15]. The easiest way to improve optical reconstruction is using a white-light source when possible. Typically, white-light sources are used in image plane configurations of classical holography [16–18].

Image plane holograms have useful property that they can be illuminated with an incoherent source (both spatially and temporally) thus reconstructing acceptably sharp images. Another advantage is increased diffraction efficiency due to the use of 1:1 beam ratios allowed since the information is coded locally [19]. Drawback is that the viewing angle is limited by the angular aperture of the imaging lens. Applying the sampling conditions to digital holography [20,21], image plane holography compared to Fresnel holography is more suitable for smaller objects. Thus, image plane digital holography is typically applied to high-speed holographic microscopy [22] and fluid velocimetry [23]. Numerically calculated image plane holograms are reconstructed by a conventional video projector with three LCOS panels [24]. For that purpose, three holograms are calculated for three wavelengths (selected according to the optimum laser reconstruction) and displayed individually by the red, green, and blue panels. Each component is thus combined and the color reconstruction achieved with a white-light source. However, the computational speed of the image plane holograms so produced [24] might be insufficient for the applications such as vibration monitoring.

Here we demonstrate the use of a white-light source for illuminating a reflective spatial light modulator (SLM) to achieve high-quality reconstructions of image-plane digital holograms. Since the raw digital holograms are usually preprocessed to remove recording imperfections and various noises, we also compare the monochrome and polychrome reconstructions for differently preprocessed holograms. The organization of the paper is as follows. In Sec. 2, the experimental system is described, where details are given for the SLM and the light sources used. In Sec. 3, the results are given for both coherent (laser) and white light (LED) illumination. Conclusions are drawn in Sec. 4.

2. Experimental setup

In this section, at first we present the main properties of the addressable SLM as well as of light sources, which are employed. A two-stage experimental setup with the possibility of recording and reconstructing image plane holograms in real-time is then described.

2.1 Spatial light modulator

For the SLM device from Holoeye Photonics AG, [type: reflective, twisted nematic, liquid crystal on silicon (LCOS); resolution: $1920H\times 1080V$, pixel pitch: 6.4 μm], first we determined experimentally the optimum polarizer/analyzer configuration (162°/95°), and then we characterized it by the standard procedure (for more details see: http://www.holoeye.com). The LCOS characterization curves show the intensity and phase modulation as a function of the gray scale level (GSL) from 0 to 255 (8-bit scale), see Fig. 1 . The intensity modulation curve, Fig. 1(a), consists of normalized reflectance values of the panel. The phase modulation curve, Fig. 1(b), is calculated from fringe shifts of the two-beam interference pattern obtained through reflection from two half’s of the total area of the panel. During measurements the GSL of one half-area was kept constant ($GSL=0$) while that of the other half-area was variable ($GSL=0,15,30,\mathrm{...},255$), thus introducing the fringe displacement as a result of the relative phase change between two beams.

 

Fig. 1 The LCOS modulation characteristics: (a) intensity and (b) phase.

Download Full Size | PDF

2.2 Light sources

The LCOS is intermittently illuminated by collimated light of two different light sources: coherent light (laser Nd Yag, single mode, $wavelength=532nm$) and one white light (LED). Their spectral characteristics are shown in Fig. 2 .

 

Fig. 2 The spectral characteristics of the sources: (a) laser (532 nm, $|\leftrightarrow |=375MHz$) and (b) white light (LED, manufacturer: Lexman).

Download Full Size | PDF

2.3 Experimental system

The experimental system consists of two parts, one for recording of the image-plane digital holograms and the other for instantaneous optical reconstruction of the captured holograms. Thus, the system functions in real-time, in which the output of the recording part (i.e. the captured holograms by photo-sensitive array detector) represents the input of the reconstructing part (i.e. the displayed holograms by the LCOS).

For recording of digital holograms, a He-Ne laser ($wavelength=633nm$) was used as light source and a monochrome CCD sensor ($1392H\times 1040V$ square pixels of 4.65 μm pitch) as light detector. For the demonstration purpose, two types of reflective objects, a static one (USAF target) and a dynamic one (two-dimensional MEMS with the mirror size equal to 1 mm²), served as input objects. The recording setup is schematically shown in Fig. 3 . As indicated in Fig. 3, the object is illuminated by a collimated laser beam almost perpendicular to the object plane. The object beam interferes with a collimated reference beam to form the digital hologram at the CCD plane. The CCD sensor is positioned at the image plane of the imaging lens and the angle between the reference and object beams is kept sufficiently small by using the cube beam splitter/combiner in front of the CCD sensor.

 

Fig. 3 Recording setup: (ND) neutral density filter, (VBS) variable beam splitter, (CBS) cube beam splitter, (CL) collimating lens, (IL) imaging lens, (SF) spatial filter, (AP) aperture, (P) polarizer.

Download Full Size | PDF

For reconstructing of digital holograms, we used two different light sources: coherent light (laser Nd-YAG, $wavelength=532nm$) and non-coherent (white light, LED). The scheme of this part of the system is shown in Fig. 4 . The light sources are aligned in a symmetric configuration by the use of the flipping (removable) mirror. Thus, all setup parameters are the same for both light sources. A color CCD sensor ($1392H\times 1040V$ square pixels of 4.65 μm pitch) is used for capturing the output hologram reconstructions.

 

Fig. 4 Reconstructing setup: (SF) spatial filter, (CL) collimating lens, (FM) flipping mirror, (ND) neutral density filter, (P) polarizer, (λ/2) half wave plate, (AP) aperture, (SAP) slit aperture.

Download Full Size | PDF

3. Results

The results of the optical hologram reconstructions are presented in Figs. 5 –7 (the USAF 1951 test target) and in two movies (MEMS, the links in Fig. 8 ), in which the monochrome (laser) reconstructions are shown on the left-hand side while the polychrome (LED) reconstructions are shown on the right-hand side of the figures. In this section, first we demonstrate the influence of the preprocessing of holograms on the monochrome and polychrome reconstructions and then the real-time performance of the system.

 

Fig. 5 Optical reconstructions of the USAF target from the single hologram: (a) laser and (b) LED.

Download Full Size | PDF

 

Fig. 6 The same as in Fig. 5, but for two subtracted single holograms.

Download Full Size | PDF

 

Fig. 7 The same as in Fig. 5, but for the reference and object recordings subtracted from the single hologram.

Download Full Size | PDF

 

Fig. 8 The real-time recordings of the two-dimensional MEMS vibrations: (a) laser reconstruction (Media 1) and (b) LED reconstruction (Media 2).

Download Full Size | PDF

3.1 Hologram preprocessing effects: coherent vs. white light

Figure 5 shows the typical results obtained by reconstructing the single hologram using the laser and LED illumination. Since subtracting of two holograms results in improved numerical reconstructions [3,25], it is of interest to investigate the corresponding optical reconstructions. As can be seen from Fig. 6 , the optical reconstructions obtained from such a hologram demonstrate apparent difference between the two types of illumination. Furthermore, as shown in Fig. 7, favorable use of the LED illumination becomes considerably more pronounced if subtracting the recordings of the reference and object wave illuminations from the single hologram recording is chosen as the preprocessing approach [26].

The subtraction method was introduced to suppress the zero-order disturbance in digital holography. The method is based on two hologram recordings of the stochastically changed wave fronts of the same object. In the preprocessing, the starting 8-bit GSL values of two holograms are first transformed into the real-valued numbers ($0\to 0.0,\mathrm{...},255\to 1.0$), then subtracted, and finally again transformed into the 8-bit GSL values. Because of performing several different renormalizations (also with negative numbers) and using the specific display mechanism, decoding of the preprocessed holograms become more sensitive to the spectral content of the light source used for optical reconstruction.

3.2 Dynamic hologram reconstructions: real-time performance

In Fig. 8, the first slides of the movies are shown demonstrating dynamic hologram reconstructions taken by the color CCD (see Fig. 4) in real-time. The movies show vibration of the two-dimensional MEMS at the resonant frequencies: (1843, 1439) Hz for the laser reconstruction, Fig. 8(a), and (1812, 1449) Hz for the LED reconstruction, Fig. 8(b).

4. Conclusion

In the present study, the real-time system is developed that uses a CCD sensor for recording and a reflective high resolution LCOS for reconstructing of the image plane digital holograms. The reconstructions obtained by using two types of light sources, namely, a coherent (laser source) and non-coherent white light (LED source) are studied and presented for comparison. It is demonstrated that a white light source can be used to achieve high-quality reconstructions of optically recorded digital holograms and holographic interferograms. Since the raw digital holograms are rarely used without preprocessing, it is also shown that the use of non-coherent light particularly improves the contrast and resolution of optical reconstructions of such preprocessed holograms. Last but not the least, it is demonstrated that the use of compact, inexpensive, low power consumption, and long life properties of LED's in digital holography could be very advantageous in designing eventual control and measuring systems.