Plenoptic cameras generally employ a microlens array positioned between the main lens and image sensor to capture three-dimensional target radiation in visible range. Because focal length of common refractive or diffractive microlenses is fixed, depth field (DOF) limited so as restrict their imaging capability. In this paper, we propose new plenoptic camera using liquid crystal (LCMLA) with electrically tunable length. The developed LCMLA fabricated by traditional photolithography standard...

An electrically tunable infrared (IR) filter based on the liquid crystal (LC) Fabry–Perot (FP) key structure, which works in wavelength range from 5.5 to 12 μm, is designed and fabricated successfully. Both planar reflective mirrors with a very high reflectivity of ∼95%, are shaped by depositing layer aluminum (Al) film over one side double-sided polished zinc selenide wafer, coupled into dual-mirror FP cavity. The LC materials filled cavity thickness ∼7.5 μm for constructing LC–FP filter,...

Real-time image lossless and near-lossless compression based on LOCO-I algorithm is in great demand many critical missions, such as satellite remote sensing, space exploration, nuclear medical imaging, for its excellent complexity/compression rate tradeoff. However, the real-time implementation of encounters two bottlenecks error prediction: context conflict update pixel reconstruction loop mode, which always limit overall throughput practice. This paper adopts a patch-wise method proposes...

An efficient two-scan connected component labelling (CCL) algorithm is proposed for a general purpose graphics processing unit (GPGPU). Compared to other GPU CCL algorithm, this has three distinct features. First, block-based and run-based strategies are combined in the first scan simplify equivalence label resolving process. Secondly, novel method introduced by constructing forest of rooted trees using only 16-bit value each node. Thirdly, whole can be implemented shared memory minimise...

In this paper, we present a kind of dual-mode photosensitive arrays (DMPAs) constructed by hybrid integration liquid crystal microlens array (LCMLA) driven electrically and CMOS sensor array, which can be used to measure both the conventional intensity images corresponding wavefronts objects. We utilize materials shape with tunable focal length. Through switching voltage signal on off, acquired through DMPAs, sequentially. use white light obtain object's for avoiding losing important...

An arrayed electrically tunable infrared (IR) filter based on the key structure of liquid crystal Fabry-Perot (LC-FP) working in wavelength range from 2.5 to 5 μm is designed and fabricated successfully. According controlled birefringence characteristics nematic LC molecules, refractive index materials filled into a prefabricated microcavity can be adjusted by spatial electric field stimulated between top bottom aluminum (Al) electrodes. As crucial component filter, Al film with typical...

A configurable 2-D adaptive digital filter architecture based on normalized least mean square (TDNLMS) algorithms is proposed. This can be used as a prewhitening to enhance the detectability of small objects in images. Every tap this configured one three types: input, reference, and disable taps. By setting type each tap, utilized implement TDNLMS prediction error any support region 5x5 array. feature allows detect whose area changes gradually during detection process, or process images with...

A fast stripe-based connected component labelling algorithm is proposed for binary image labelling. Stripe extraction strategy used to transform the pixel-connected issues, which most of previously algorithms focused on, into stripe-connected issues. The stripe-union treats combination neighbouring stripes as mergence rooted trees. Finally, comparisons are performed with other famous algorithms. has shown better performance than state-of-the-art in real images test, while auxiliary memory...

Determination of relative heights from SEM stereo photographs is considered. It shown that by the choice suitable geometry it possible to derive exact formulas for both case translating and tilting specimen. The translation method be useful determining true working distance whereas more sensitive small height differences. A comparison between approximate (infinite distance) made in specimen case. Experimental requirements are considered methods suggested.

We present a new liquid crystal (LC) microlens structure having ommatidia function, which consists of the overlapped patterned electrode layers. Each layer has circular array with different size. Two layers are deposited on surface glass substrate and insulated by thin SiO₂ coating, act as controlling electrodes. A plane another acts base electrode. substrates made into LC cell filled nematic material. When voltage signal is applied to electrode, each unit can focus along optical axis good...

A Support Vector Machine (SVM) based method for land and sea segmentation in Polarimetric SAR (POLSAR) is proposed this paper. The principle of SVM first briefly summarized. Features that selected consist 9 polarimetric features obtained from target decompositions, i.e., Krogager, Freeman-Durden Cloude 6 texture calculated first-order statistics. These 15 are combined to feature vectors. experiments carried out on POLSAR data Radarsat-2. classifier through training with samples then applied...

In Cartesian coordinate systems, the angular separation-based star identification algorithms involve much trigon- ometric function computing. That delays algorithm process. As in a polar system, coordinates are denoted by values, it is potential to speed up process adopting sys-tem. An system introduced and grid based on proposed enhance performances of The simulations demonstrate that superior rectangle com-puting cost rate. It can be used sensors for high precision reliability spacecraft...

SIFT (Scale Invariant Feature Transform) is one of most popular approach for feature detection and matching. Many parallelized algorithms have been proposed to accelerate apply into real-time systems. This paper divides the researches three different categories, that is, optimizing parallel based on general purpose multi-core processors, designing customized processor dedicated implementing FPGA (Field Programmable Gate Arrays). Overview type analysis task-level parallelism are presented in...

Metallic electric split-ring resonators (SRRs) with featured size in micrometer scale, which are connected by thin metal wires, patterned to form a periodically distributed planar array. The arrayed metallic SRRs fabricated on an n-doped gallium arsenide (n-GaAs) layer grown directly over semi-insulating (SI-GaAs) wafer. microstructures and n-GaAs construct Schottky diode, can support external voltage applied modify the device properties. developed architectures present typical functional...

A low‐cost memory data scheduling method based on two N /2‐depth single‐port memories is proposed for reconfigurable fast Fourier transform (FFT) bit‐reversed reordering tasks. To make have the equivalent ability to read and write simultaneously, types of address generation methods are proposed. Based method, bit‐reversal circuits designed continuous The design implemented a maximum 8 k flexible length FFT processor. Compared with other conventional methods, can reduce area cost by 53.8...

A dual-film metamaterial is proposed that presents oscillation characteristics during terahertz transmission. An n-type gallium arsenide (n-GaAs) film with a free electron density of 4.7 × 1017 cm−3 grown on semi-insulating GaAs wafer. Then, metallic film, which fabricated the n-GaAs patterned into an arrayed single-gap microstructure narrowed edges using traditional ultraviolet photolithography methods. The metal and films form Schottky junction. transmission frequency spectrum contains...

针对基于场景的自适应校正算法普遍存在鬼影的问题, 分析了神经网络算法(NN-NUC)产生鬼影的原因,并在此基础上提出了用基于偏微分方程(PDE)的非线性滤波方法取代NN-NUC算法中邻域平均的方法来获取期望图像,从而减少边缘像素误差,达到消除鬼影的目的.采用实际采集的红外图像进行实验,结果表明,很好地消除了鬼影.与已有的几种去鬼影的方法相比,具有更快的收敛性.;Generally, most of adaptive nonuniformity correction algorithms have the ghosting artifact problem. In this paper, cause artifacts in Neural Network (NN-NUC) algorithm for infrared focal plane array (IRFPA) was studied. Based on analysis, a novel eliminating proposed, which replaces linear spatial average filter...

Based on our previous works liquid crystal (LC) microlenses driven electrically, we present a new type of dual-mode microlens arrays (DLCMAs) for chaotic encryption applications. Currently, the DLCMAs developed by us consist top electrode couple constructed two layers controlling and bottom planar electrode. Aluminium Indium-Tin Oxide (ITO) materials are respectively deposited over both sides glass substrate shaping couple, which is used to act as key mode-control-part in DLCMAs. Another ITO...

The paper investigates a binarized backbone neural network design optimization method for dedicated hardware circuits, in which the activation values and weights of convolutional operations are represented by 1bit to achieve low computation number parameters while maintaining high image classification accuracy. adopts hyperparameterized training method, introduces hyperparameters binarization process function, improve accuracy, inference fuses all absorption into two threshold numbers...

An efficient connected component labeling algorithm for multi-value image is proposed in this paper. The simple and inerratic suitable hardware design. A one-dimensional array used to store equivalence pairs. Record organization of table advantageously find the minimum equivalent label, can shrink time on processing table. pipelined architecture described enhance system performance.