Open Access
Add to BibSonomyAbstract
A novel fiber reference optical readout method was proposed in the bi-material micro cantilever infrared imaging system, which consists of an infrared imaging channel, an optical readout channel and a fiber reference channel. The fiber reference channel is used to monitor the intensity fluctuation of the light source, and provide a signal to correct the distortion of the infrared images from the optical readout channel. Comparing with the typical optical readout method without any references, the noise equivalent temperature difference (NETD) of such an infrared imaging system with the fiber reference optical readout method can be reduced by about 33% and edges of the IR images become clearer.
©2012 Optical Society of America
1. Introduction
In 1996, Thomas Thundat in Oak Ridge National Lab invented a new infrared (IR) detector with voltage resistance micro cantilever beam, which showed the feasibility of using bi-material cantilever as infrared detector [1]. Therefrom many research institutions began to study the IR imaging system using the bi-material micro cantilever detector. Comparing with the thermoelectric-typed uncooled infrared imaging system, it eliminates the need for complex readout circuit, which makes it have extensive applied perspective [2,3].
The NETD of the optical readout infrared FPA imaging system can be theoretically reduced to less than 5.7mK [4] or 5mK [5], however, it is usually almost 120mK [6] or more in practice, which may be due to the fact that there are many factors could degrade the image quality, such as the undesired shape distortions in fabrication of the FPA, the aberrations of the lens, the impact of the filter, the stability of the illumination, and so on. In 2007, an optical readout platform using a knife-edge filter for detecting the bending of a bi-material micro cantilever array was established by Miao et al, and the NETD could be reduced to 50% [7]. In 2007, Liu et al pointed out that using holography illumination to compensate the shape distortions of the FPA in the optical readout infrared FPA imaging system was a feasible way to improve the system performance [8].
The former researchers mainly focused on the improvement of the filter and the compensation of the undesired shape distortions during fabrication of the FPA. In practice, the output of IR images are based on the gray-level difference, while the luminous intensity of the light source changes with time, which affects the image quality greatly. Generally, LED is often used as the light source and its luminous intensity alters owing to the fluctuations of ambient temperature, the electric current or other factors. To solve the problem mentioned above, this paper presents a new method which uses a fiber reference channel to monitor the fluctuation of the light source. The signal from the fiber reference channel is used to revise the images in real time, which can thus improve the NETD of the system.
It has been validated by means of experiment that the proposed method is effective in increasing the sensitivity of the proposed IR imaging system and improving its images quality.
2. Methods and instruments
2.1 Schematic diagram
The novel optical readout infrared FPA imaging system with fiber reference channel is a dual-light path system, as shown in Fig. 1 .
The system was composed of the IR imaging channel, the optical readout channel and the fiber reference channel. Comparing with the typical optical readout infrared FPA imaging system, the key point of the current system was the accretion of the fiber reference channel.
The IR lens collects the IR radiation of the object by focusing them on the FPA. For every micro cantilever pixel in the FPA, its absorber/reflector structure has one layer as an infrared absorber made by SiNx, and the other layer as a visible light reflector made by Au. Each pixel’s micro cantilevers of the FPA don’t deform when there is no infrared radiation focused on it, but its temperature rises after absorbing the infrared radiation, and the cantilevers deflect due to the difference between the thermal expansion coefficients of the two layer materials. For each pixel, the deflected angle θ is proportional to the incident IR radiation focused on it, namely, different IR radiations from different object points will make different deflected angle θ. The optical readout channel detects the deformation of the cantilevers of the FPA and the fiber reference channel monitors the intensity fluctuation of the light source and corrects its influence on the image.
In the optical readout channel, the light emitted from the light source is collimated by its illumination lens, and divided into two beams by a beam splitting prism. One of the beams is directed to the FPA for optical imaging, and the other is focused into the fiber for the reference. The rays reflected by the FPA are focused on the filter plane after passing through the prism again and the Fourier lens, then the filtered beam is projected onto CCD by the imaging lens. In the fiber reference channel, the reference rays emitted from the fiber directly illuminate a sub-area of CCD, pixels of which are divided into two areas. The great majority pixels of CCD receive the IR imaging from the optical readout channel, and the certain pixels receive the reference signals from the fiber reference channel, as shown in Fig. 2 .
The energy that the great majority pixels of CCD receive is Ea when there is no deformation on the micro cantilever of the FPA. It will change to Es when the cantilever of the FPA deflects due to the object IR radiation. Then the energy variation that the great majority pixels of CCD detect is ∆E = Ea-Es under the two conditions. The image processing system displays ∆E on the monitor in the form of gray-level difference.
∆E is always caused by not only the object IR radiation but also other factors, such as the fluctuation of the light source intensity. In order to get the energy variation caused by the object IR radiation only, the energy variation of the fluctuation of the light source intensity must be deducted. The fiber reference channel is used to monitor the intensity fluctuation of the light source and provides a signal to deduce the effect of the fluctuation of the light source on the ∆E and improve the image quality of the system.
2.2 Principles
From the analysis above, we know that the display image is a gray-level image, and the gray-level is proportional to the energy difference ∆E that each pixel receives, giving
where Ea is the energy that the IR image area of CCD received when there’s no object infrared radiation, and Es is the energy when there’s object infrared radiation.In practice, the object IR radiation is blocked at the time t0 and the energy that the IR image area of CCD receives is E0. The object IR radiation is let in at the time t and the energy that the IR image area of CCD receives is Et. Substituting Et and E0 for Es and Ea, we get
The light source of this system is a green LED and its luminous intensity always changes when the ambient temperature, the electric current or any other factors alter, which leads to the variation between ∆E and ∆E0, as shown in Fig. 3 ,
∆Er in Fig. 3 is the real IR image caused by the object IR radiation, while ∆Ee is generated by the fluctuation of the light source. If ∆E0 is used as the IR image, the IR image will be amplified or neutralized owing to the existence of ∆Ee, which leads to the anamorphose. So we must get ∆Er from ∆E0. The fiber reference channel is used to achieve this goal.
Supposing that the energy from the fiber reference channel is Eq0 at the time t0 and Eqt at the time t, when the object IR radiation is blocked, the following equation can be obtained
Where Eat is the energy from the optical readout channel at the time t, when the object IR radiation is blocked.Since Eq0 and Eqt can be gained directly from the fiber reference channel and E0 can be gained directly from the optical readout channel, according to Eq. (3), one can get
From Eq. (1), the real gray-level value at the time t can be expressed as
Substituting Eq. (4) into Eq. (5), we get
Normalizing ∆E according to the time t0, leads to
Compared Eq. (7) with Eq. (2), it can be seen the latter one has an extra modifying factor (Eq0/Eqt). The factor can be used to revise the background signal at the time t to the time t0, in order to eliminate the effect caused by the energy difference between the time t0 and t due to the light source luminous intensity changes, such as ∆Ee in Fig. 3.
3. Experimental results
The infrared imaging system was set up according to Fig. 1, and we got infrared images shown in Fig. 4 and Fig. 5 .
Figure 4 shows the images of an electric iron three meters away. Figure 4(a) is calculated on the basis of Eq. (2) and Fig. 4(b) is calculated on the basis of Eq. (7).
Figure 5 shows the image of a scissors three meters away in front of the background at 70°C . Figure 5(a) is calculated on the basis of Eq. (2) and Fig. 5(b) is calculated on the basis of Eq. (7).
From Fig. 4 and Fig. 5, it can be seen that the images revised by the fiber reference method have better contrast and clearer outline than the uncorrected ones.
3.1 Sensitivity
The NETD represents the detection sensitivity of the IR imaging system, which is one of the key qualifications of the IR imaging system. It can be expressed as
where ∆T is the temperature difference, Vs is the peak voltage of the signal, and Vn is the RMS value of the noise.In this system, the peak voltage of the signal Vs is in direct ratio with the corresponding gray-level value of the image Gmax. The RMS voltage of the noise Vn is proportional to the RMS offset of the noise gray-level value Gn. Thus, substituting Gmax and Gn for Vs and Vn, leads to
where k is a coefficient.By picking the parameters from Fig. 4(a) and Fig. 4(b), the corresponding value of NETD can be obtained, as shown in Table 1 , from which it can be seen that the NETD values was reduced about 33% by using the fiber reference method.
Table 1. Calculated Values of NETD
3.2 Image definition
The original angles θ0 of the different micro cantilever of the FPA aren’t entirely consistent owing to the processing technique. Thus, the Fraunhofer diffraction pattern of the different unit of the FPA on the spectrum plane doesn’t overlap completely, as shown in Fig. 6 .
Since the reflected light of each FPA unit that enters into the filter hole isn’t the same, the impact of ∆Ee on ∆Er is different when the light source intensity changes, as shown in Fig. 7 .
We can see that impact of ∆Ee on ∆Er in Fig. 7(b) is much more than that in Fig. 7(a). Therefore, even if the object IR radiation is the same, the light source intensity fluctuation will held a greater influence on the Fig. 7(b) than the Fig. 7(a), which results to the distortion of the energy variation on the imaging plane and the edges of the images will inevitably appear distorted and fuzzy. The adoption of the fiber reference channel eliminates the impact of the light fluctuation ∆Ee, thus, the edges of the IR images become clearer and the gray-level values are more uniform such as in Fig. 5.
4. Conclusions
In summary, a new approach named “the novel optical readout infrared FPA imaging system with fiber reference channel” was presented. Comparing with the conventional optical readout method, this method can effectively eliminate the influence of the light fluctuation and other miscellaneous light, increase the sensitivity of the system and improves the image quality. The NETD can be reduced by 33%, and the edges of the IR images become clearer and the gray-level value is more uniform.
Acknowledgments
This work was supported by the State Key Program of National Natural Science Foundation of China (No. 61036006), the National Natural Science Foundation of China (No. 61177094) and the Doctoral Program Foundation of Institutions of Higher Education of China (No. 20091101120022). The authors would like to thank Dr. Jing Sui from The Mind Research Network for her great suggestion to the paper.
References and links
1. P. I. Oden, P. G. Datskos, T. Thundat, and R. J. Warmack, “Uncooled thermal imaging using a piezoresistive microcantilever,” Appl. Phys. Lett. 69(21), 3277–3279 (1996). [CrossRef]
2. J. Varesi, J. Lai, T. Perazzo, Z. Shi, and A. Majumdar, “Photothermal measurements at picowatt resolution using uncooled micro-optomechanical sensors,” Appl. Phys. Lett. 71(3), 306–308 (1997). [CrossRef]
3. A. Rogalski, “Infrared detectors: Status and trends,” Prog. Quantum Electron. 27(2-3), 59–210 (2003). [CrossRef]
4. J. Zhao, “High Sensitivity Photomechanical MW-LWIR Imaging using an Uncooled MEMS Microcantilever Array and Optical Readout,” Proc. SPIE 5783, 506–513 (2005). [CrossRef]
5. H. Torun and H. Urey, “Uncooled Thermo-mechanical Detector Array with Optical Readout,” Proc. SPIE 5957, 59570O, 59570O-9 (2005). [CrossRef]
6. J. P. Salerno, “High Frame Rate Imaging Using Uncooled Optical Readout Photomechanical IR Sensor,” Proc. SPIE 6542, 65421D, 65421D-9 (2007). [CrossRef]
7. Z. Miao, Q. Zhang, Z. Guo, X. Wu, and D. Chen, “optical readout method for microcantilever array sensing and its sensitivity analysis,” Opt. Lett. 32(6), 594–596 (2007). [CrossRef] [PubMed]
8. M. Liu, Y. Zhao, L. Dong, X. Yu, X. Liu, M. Hui, J. You, and Y. Yi, “Holographic illumination in optical readout focal plane array infrared imaging system,” Opt. Lett. 34(22), 3547–3549 (2009). [CrossRef] [PubMed]
