We have fabricated organic thin-film transistor (OTFT)-driven active matrix liquid crystal displays on flexible polymeric substrates. These small 16×16 pixel polymer-dispersed arrays addressed by pentacene layer OTFTs. The were using a low-temperature process (<110 °C) polyethylene naphthalate film and are operated as reflective displays.

We report on parasitic contact effects in organic thin film transistors (OTFTs) fabricated with pentacene films. The influence the OTFT performance of source and drain metal device design was investigated. Top (TC) bottom (BC) gated transmission line model (gated-TLM) test structures were used to extract combined resistance as a function gate voltage swing drain-source for OTFTs gold contacts. For comparison BC palladium contacts studied. Differences bias dependence TC indicate that charge...

In this paper, we address the problem of extrinsic calibration a camera and 3D Light Detection Ranging (LiDAR) sensor using checkerboard. Unlike previous works which require at least three checkerboard poses, our algorithm reduces minimal number poses to one by combining line plane correspondences. Besides, prove that parallel planar targets with boundaries provide same constraints in algorithm. This allows us place close LiDAR so laser points better approximate target boundary without loss...

We report on organic thin-film transistors fabricated using the small-molecule semiconductor naphthacene as active layer material with device performance suitable for several large-area or low-cost electronics applications. investigated thin films deposited by thermal evaporation onto amorphous substrates held near room temperature. Using atomic-force microscopy and x-ray diffraction we find consist of a high density submicron-sized grains surprisingly degree molecular order. Thin-film...

This paper presents a new algorithm to automatically solve algebra word problems.Our solves problem via analyzing hypothesis space containing all possible equation systems generated by assigning the numbers in into set of system templates extracted from training data.To obtain robust decision surface, we train log-linear model make margin between correct assignments and false ones as large possible.This results quadratic programming (QP) which can be efficiently solved.Experimental show that...

Planes ubiquitously exist in the indoor environment. This letter presents a real-time and low-drift LiDAR SLAM system using planes as landmark for Our algorithm includes three components: localization, local mapping global mapping. The localization component performs registration, instead of scan-to-scan registration adopted state-of-the-art odometry (LOAM) framework that yields lower fidelity poses. optimizes poses keyframes within sliding window parameters observed by these keyframes....

We use grazing incidence x-ray scattering to study the molecular structure and morphology of thin (70 ML) crystalline films 3,4,9,10-perylenetetracarboxylic dianhydride (PTCDA) on Au(111) surfaces as a function film thickness, substrate temperature, growth rate. Although first two PTCDA monolayers grow in layer-by-layer fashion, evolution beyond second monolayer depends strongly upon conditions resulting low-temperature [i.e., nonequilibrium (NEQ)] high-temperature [equilibrium (EQ)]...

In this paper, we present a new minimal solution for the extrinsic calibration of 2D light detection and ranging (LIDAR) sensor perspective camera. This problem is formulated as registering three planes corresponding coplanar lines. All existing algorithms solve by its geometric structure in original or dual 3D space. contrast, our algorithm directly exploits algebraic polynomial system to resolve problem. more abstract, however, provides broadly applicable method other problems that need...

Real-time traffic sign detection and recognition are essential challenging tasks for intelligent vehicle. The previous works mainly focus on detecting recognizing based images captured by onboard camera. Visual features of such as color, shape, appearance, however, often sensitive to illumination condition, angle view, etc. Except camera, LIDAR also provides important alternative sign. fusion complementary data acquired from both sensors can improve the robustness algorithm, especially when...

In this paper, we propose a novel robust algorithm for the extrinsic calibration of camera and lidar. This utilizes checkerboard as object. Since interaction between estimation errors plane parameters obtained from images downgrades quality results, new geometric constraint is presented to decouple rotation translation so reduce effect such an interaction. Weights that represent uncertainty unit normal vector are introduced totally evaluate each pair image lidar scan. Furthermore, analyze...

This paper introduces a real-time dense planar LiDAR SLAM system, named π-LSAM, for the indoor environment. The widely used odometry and mapping (LOAM) framework [1] does not include bundle adjustment (BA) generates low fidelity tracking pose. seeks to overcome these drawbacks Specifically, we use plane as landmark, introduce (PA) our back-end jointly optimize planes keyframe poses. We present π-factor significantly reduce computational complexity of PA. In addition, an efficient loop...

Planes, lines, and cylinders widely exist in man-made environments. This letter introduces a LiDAR simultaneous localization mapping (SLAM) system using those three types of landmarks. Our algorithm has components including local mapping, global localization. The jointly adjust planes, with poses to minimize the point-to-model cost, which is referred as plane-line-cylinder adjustment (PLCA). We prove that, some preprocessing, PLCA independent number points captured from landmarks, makes...

This paper introduces an efficient direct visual odometry (VO) algorithm using points and lines. Pixels on lines are generally adopted in methods. However, the original photometric error is only defined for points. It seems difficult to extend it In previous works, collinear constraints either ignored [1] or introduce heavy computational load into resulting optimization system [2]. extends We prove that 3D of a 2D line determined by inverse depths endpoints line, derive closed-form solution...

In this letter, we present a direct visual odometry (VO) using points and lines. Direct methods generally choose pixels with sufficient gradients to minimize the photometric error for status estimation. Pixels on lines are involved in process. But collinear constraint among these ignored, which may result less accurate depth This letter introduces into state-of-the-art DSO (J. EngelI et al., 2018) overcome problem. The 3D lines, poses within sliding window jointly optimized. Points only...

In order to constrain the evolutionary state of red supergiant Betelgeuse, we have produced a suite models with ZAMS masses from 15 25 Msun in intervals 1 including effects rotation. The were computed stellar code MESA. For non-rotating find results that are similar other work. It is somewhat difficult agree within sigma observed values R, Teff and L, but modestly easy 3 uncertainty. Incorporating nominal rotational velocity, ~15 km/s, yields significantly different, challenging,...

This paper presents an efficient algorithm for the least-squares problem using point-to-plane cost, which aims to jointly optimize depth sensor poses and plane parameters 3D reconstruction. We call this Planar Bundle Adjustment (PBA), due similarity between original (BA) in visual As planes ubiquitously exist man-made environment, they are generally used as landmarks SLAM algorithms various sensors. PBA is important reduce drift improve quality of map. However, directly adopting...

Voxel-based methods have achieved state-of-the-art performance for 3D object detection in autonomous driving. However, their significant computational and memory costs pose a challenge application to resource-constrained vehicles. One reason this high resource consumption is the presence of large number redundant background points Lidar point clouds, resulting spatial redundancy both voxel BEV map representations. To address issue, we propose an adaptive inference framework called Ada3D,...

We present a novel unsupervised learning framework for single view depth estimation using monocular videos. It is well known in 3D vision that enlarging the baseline can increase accuracy, and jointly optimizing set of camera poses landmarks essential. In previous frameworks, only part photometric geometric constraints within sequence are used as supervisory signals. This may result short overfitting. Besides, works generally estimate low resolution from impute image. The then interpolated...

Road curb detection and tracking is essential for the autonomous driving of intelligent vehicles on highways urban roads. In this paper, we present a fast robust road algorithm using 3D lidar data Integral Laser Points (ILP) features. Range intensity decomposed into elevation projected ground plane. First, left right curbs are detected each scan line range segment Then, points determined data. The ILP features proposed to speed up both procedures. Finally, parabola model RANSAC used fit...

In this paper, we propose a new extrinsic calibration algorithm for camera and 2D LIght Detection And Ranging sensor (LIDAR). N plane-line correspondences are used to establish geometric constraints computing the parameters. The existing algorithms that use same either require ≥ 5 or focus on minimal solution = 3. contrast, proposed is effective all This made possible by decoupling initial estimation of rotation from translation imposing orthonormal computation rotation. When matrix...

Reliable detection and recognition of planar objects including traffic sign, street road surface in dynamic cluttered natural scenes are a big challenge for self-driving cars. In this paper, we propose comprehensive method object recognition. First, the data association LIDAR camera is set up to acquire colorized laser scans, which simultaneously contain both color geometrical information. Second, combine three spaces RGB, HSV, CIE L*a*b* with reflectivity as an aggregation-based feature...

This letter presents a self-supervised framework for learning depth from monocular videos. In particular, the main contributions of this include: (1) We present windowed bundle adjustment to train network. Compared most previous works that only consider constraints consecutive frames, our increases camera baseline and introduces more avoid overfitting. (2) extend widely used U-Net architecture by applying Spatial Pyramid Net (SPN) Super Resolution (SRN). The SPN fuses information an image...

In this paper, we address the problem of pose estimation from N 2D/3D point correspondences, known as Perspective-n-Point (PnP) problem. Although many solutions have been proposed, it is hard to optimize both computational complexity and accuracy at same time. propose an accurate simultaneously efficient solution PnP Previous algorithms generally involve two sets unknowns including depth each pixel camera. Our formulation does not pixel. By introducing some intermediate variables, leads a...

This paper proposes a novel algorithm to solve the pose estimation problem from 2D/3D line correspondences, known as Perspective-n-Line (PnL) problem. It is widely that minimizing geometric distance generally results in more accurate than an algebraic distance. However, rational form of reprojection yields complicated cost function, which makes solving first-order optimality conditions infeasible. Furthermore, iterative algorithms based on are time-consuming for large-scale In contrast...