Imaging with two-axis micromirrors

We demonstrate a means of creating a digital image by using a two axis tilt micromirror to scan a scene. For each different orientation we extract a single grayscale value from the mirror and combine them to form a single composite image. This allows one to choose the distribution of the samples, and so in principle a variable resolution image could be created. We demonstrate this ability to control resolution by constructing a voltage table that compensates for the non-linear response of the mirrors to the applied voltage. Micro-Opto-Electro-Mechanical Systems (MOEMS) is a relatively new field which appears to have a multitude of applications. Perhaps the best known MOEMS is the Texas Instruments’ Digital Micromirror Device (DMD), which is an NxN array of SRAM cells, each covered by a tilting mirror [1]. Each of these mirrors is either in a binary “on” or “off” state. The primary application of this device is for the projection of images. Optical switching is another application area for MOEMS. Here, mirrors can eliminate the costly conversion from the optical domain to the electrical domain for switching. An example is Lucent’s WaveStar LambdaRouter, which used an 8x8 array of 2-axis 600 μm diameter micromirrors to achieve fiber array switching for 256 channels [2]. A later version of the array consisted of 16x16 mirrors. Each mirror may be individually actuated and can achieve 100,000 distinct states [3]. In this letter, we present our initial results on photographic imaging with micromirror arrays. Using a prototype which employs a 16x16 Lucent micromirror array we have been able to obtain low resolution images with a single mirror, and illustrate a means of calibrating the mirrors for imaging purposes. Note that while the term “imaging” is commonly associated with MOEMS, it is usually not used in the sense of photographic imaging [4]. Exceptions include work done by Nayar et al. with the DMD to extend dynamic range and work by Last et al. on micro-cameras which use a 1-axis mirror to increase the field of view[5, 6]. Previous work by the authors described simulations and some manual experiments[7]. In conventional macroscopic photography there is the notion of image mosaicing, in which one combines two or more images with some common overlap to create a single image of higher