Search

In this paper, we present the design of an interferometric structure based on a planar waveguides 3D network, able to optically perform the wavelet decomposition of two dimensional input data, such as image pixels. The possibility to implement a system that could offer potential in terms of integrability and robustness is explored. This idea is supported by the existence of methods and optical architectures, based on interferometric structures, for the implementation of the optical wavelet transform. They enable parallel processing on one dimension of the input data and they can be enhanced with the development of structures able to parallelize the processing on two dimensional data. [ABSTRACT FROM PUBLISHER]

The use of phase has a long standing history in optical image processing, with early milestones being in the field of pattern recognition, such as VanderLugt's practical construction technique for matched filters, and (implicitly) Goodman's joint Fourier transform correlator. In recent years, the flexibility afforded by phase-only spatial light modulators and digital holography, for example, has enabled many processing techniques based on the explicit encoding and decoding of phase. One application area concerns efficient numerical computations. Pushing phase measurement to its physical limits, designs employing the physical properties of phase have ranged from the sensible to the wonderful, in some cases making computationally easy problems easier to solve and in other cases addressing mathematics' most challenging computationally hard problems. Another application area is optical image encryption, in which, typically, a phase mask modulates the fractional Fourier transformed coefficients of a perturbed input image, and the phase of the inverse transform is then sensed as the encrypted image. The inherent linearity that makes the system so elegant mitigates against its use as an effective encryption technique, but we show how a combination of optical and digital techniques can restore confidence in that security. We conclude with the concept of digital hologram image processing, and applications of same that are uniquely suited to optical implementation, where the processing, recognition, or encryption step operates on full field information, such as that emanating from a coherently illuminated real-world three-dimensional object. [ABSTRACT FROM AUTHOR]

We propose two methods of fabric surface imaging and processing so that the information of wrinkles can be extracted and evaluated. In the first method, we consider fringe projection and Fourier transform profilometry. In the second method we consider two images of the same sample obtained under different lateral illumination and apply a joint edge detector to integrate the information about wrinkles from both of them. The smoothed perception of wrinkles with distance and its influence in the evaluation of the fabric appearance is also studied. [ABSTRACT FROM AUTHOR]

We introduce the gyrator transform which produces a rotation of two-dimensional signal in phase space. Main properties of this transform are considered. The flexible optical set up for gyrator transform experimental implementation is discussed. The applications of the gyrator transform to image processing such as filtering, noise reduction, encryption are demonstrated. [ABSTRACT FROM AUTHOR]