e., a CMOS imager [1]. The integration of an increased number of camera functions onto a single silicon chip offers nevertheless significant advantages in terms of system miniaturization and manufacturing cost [1]. As a result, CMOS imagers meantime can now be found in a wide range of consumer electronic products from mobile phones, PC mice and webcams to fax machines, to name a few. Other examples of applications of CMOS imagers include blind spot detection and rear vision in cars, automated industrial product quality Inhibitors,Modulators,Libraries inspection, photography, digital radiography, fluoroscopy, microscopy and even implantable microimager retinal stimulation [2]. CMOS camera manufacturers have been continuously improving the performance of their products in terms of resolution, power consumption, and read-out speed [1].

The continuous aggressive scaling of the minimum feature size Inhibitors,Modulators,Libraries in CMOS technology offers the possibility to further miniaturize CMOS Inhibitors,Modulators,Libraries imagers, while still integrating increased built-in functionalities and advanced on-chip Inhibitors,Modulators,Libraries processing for improved imaging performance [1]. However, a further miniaturization of the optics is not possible since this would cause light to pass through a very small opening and to diffract or spread due to the interference of light waves. Diffraction would cause so much spreading that the image would be very Inhibitors,Modulators,Libraries blurry and essentially useless. To achieve a thin camera-on-a chip, Tanida et al have recently proposed an imaging system called TOMBO [3] (an acronym for Thin Observation Module by Bound Optics) emulating the visual system of insects and other arthropods [4, 5].

In contrast Inhibitors,Modulators,Libraries to traditional camera systems, TOMBO is not Inhibitors,Modulators,Libraries based on a single lens system [3]. Instead, a TOMBO imager comprises a collection of imaging units, each of which consists of a microlens unit associated to a subset of the pixel array (Fig. Inhibitors,Modulators,Libraries 1). Drug_discovery Adjacent imaging units are separated by an opaque wall to prevent crosstalk. Each individual imaging unit is thus optically isolated and images part of the scene. As a result, the TOMBO sensor captures multiple low resolution (LR) images at the same time and the output is a compound image formed by the mosaic of low resolution unit images.

Advanced Digital signal processing can then be used to reconstruct a high resolution image from the available set of low resolution images [3, 10].

Tanida et al have demonstrated that the restoration of a high resolution image is possible since each imaging unit provides a different view of the scene [7]-[12]. In [3], Tanida Anacetrapib et al proposed an image restoration algorithm which uses the back-projection INCB-018424 free overnight delivery (BP) method. This algorithm requires complete knowledge of the imaging system point spread function (PSF). To reconstruct the original image, the inverse (pseudo-inverse) of the known PSF is multiplied by the captured low resolution images.