Search

Dynamic microscale surface topographies are desired in smart optics, controlling surface wettability and preventing marine biofouling. Voltage-controlled reversible responses have demonstrated potential for reliable reproducibility and stability, fast response, and for actuating thin films fixed over large solid surfaces. To obtain reversible deformation with regular geometric patterns, however, electrical methods have had to be coupled with mechanical stretching/bending and other ways to induce anisotropy. There is a great need and potential for on-demand electrical generation of programmable complex responsive surface patterns. Here we demonstrate a responsive polymer coating over an underlying pattern of counter electrodes which can be activated selectively. We present a patternable electrode printing method to achieve localized and structured wrinkling deformation without mechanical pre-force deformation. We discover that below a minimal separation distance, electrodes below the polymer act as a single electrode. We establish parameters that govern the alignment of wrinkles and quantify the regularity and direction of the new patterns between electrodes separated by larger than this distance. We analyze and quantify the regularity of the formed wrinkling patterns by four disorder metrics: box-counting fractal dimension, tortuosity, angle distribution, and branch number. We demonstrate the application of such electrode/wrinkles-on-demand patterning with a working multi-state light reflection–diffusion–grating device.

We address the problem of face recognition from image sets, where subject-specific subspaces instead of image vectors are compared. Previous methods based on Grassmannian subspace distances mainly take linear subspaces as input. The non-linearity exists when the input data contain complex structure such as pose changes. We generalize Grassmannian distances into high dimensional feature space with kernel trick to handle the underlying non-linearity in data. We show that kernel Grassmannian distances in feature space can be implicitly computed from the input data. Furthermore, we propose to use projection kernel in feature space for discriminant analysis. Comparisons with several state-of-the-art methods were performed on two databases, CMU PIE and YaleB. The proposed methods have demonstrated promising performance.