Subpixel accuracy for extracting groove center based on corner detection

0IntroductionWelds quality is directly affected by the location accu-racy of welding torch to groove center in vision sensor robotwelding. Therefore, it is very significant to increase theextraction accuracy of groove center. A common method isto use high resolution CCD and high speed image collectingcard, but it costs high or sometimes there are limits inuse.Subpixel detection is an effective soft method to in-crease image location accuracy[1 -2]. Compared with chan-ging hardwares , it is v…     

CCD噪声模型在几何最优线性姿态估计中的应用

星敏感器噪声主要来源于CCD图像传感器和控制与数据处理电路,这些噪声对姿态估计精度的影响不容忽视.在研究了CCD图像传感器中光子散粒噪声、暗电流噪声、电荷转移噪声及读出噪声四种主要噪声的基础上,分析了它们对亚像元质心法提取精度的影响.根据亚像元质心法提取结果,采用几何解析的最优线性姿态估计(OLAE)方法解算卫星姿态.仿真结果表明:考虑星敏感器噪声模型比单纯引入高斯白噪声具有更高的姿态估计精度.     

Shadow geometry maps for alias-free shadows

Shadow maps sample scene visibility in the light source space and offer an efficient solution to generate hard shadows.However,they suffer from aliasing artifacts because of discretization errors,inadequate resolution and projection distortion.In this paper,we propose the shadow geometry map method,where a shadow depth map is augmented by storing geometry information about scenes.This leads to a new shadowrendering algorithm that combines a supersampling filter,a geometry-aware reconstruction kernel and an irregular sampling filter.Our method produces high quality alias-free and subpixel supersampling shadow rendering and retains the simplicity and the efficiency of shadow maps.We show that the algorithm pipeline is efficiently parallelized using current programmable graphics hardware and that our method is capable of generating high quality hard shadows.     

局部相位相关用于图像亚像素级配准技术研究

提出了一种基于局部相位相关的高效和鲁棒的亚像素级图像配准方法。通过传统的相位相关算法估计出初始平移参数后,在初始位置的引导下对互相关功率谱进行上采样矩阵Fourier变换,实现了图像局部相位相关,得到图像间亚像素级平移参数。实验结果表明,算法配准精度较高,且对随机噪声和光照变化具有较强的鲁棒性。     

基于小波分析的亚像素配准算法

在像素级配准的基础上,通过对多项式细分算法的改进,提出了一种基于小波分析的亚像素配准算法,并对配准算法的精度进行了分析研究.仿真结果表明,该算法效果良好,达到亚像素级精度.     

基于约束空间光谱联合的亚像素定位方法北大核心CSCD

针对高光谱遥感图像,提出了一种约束空间光谱的亚像素定位方法。传统的亚像素定位方法以解混的结果作为输入,可能无法充分利用高光谱图像丰富的光谱信息。本文所提出的基于约束空间光谱联合的亚像素定位方法(constraint spatial-spectral subpixel mapping,CSSSM),利用下采样将像素丰度与亚像素丰度显式联系起来,代入线性解混模型得到亚像素丰度求解的新模型。在求解过程中,通过添加稀疏性约束与平滑性约束,以限制亚像素丰度的解空间,亚像素丰度求解更精确。其中,针对亚像素丰度稀疏性先验采用重加权1范数作为新的约束,并自适应地更新权重;针对亚像素丰度空间先验信息则采用全变分(total variational,TV)正则化作为约束,然后使用乘法迭代算法求解亚像素丰度,最后利用赢者通吃的策略进行类别确定。在两个合成数据集上进行了实验,结果表明,本方法能够进一步提高亚像素定位的精度。     

一种AVS亚像素运动估计快速算法

为提高图像的质量和压缩率,AVS视频标准中的运动补偿精确到了1/4像素,但采用亚像素全搜索算法的计算量太大,导致整个编码器的编码速率很低。针对该问题,提出一种新的亚像素运动估计快速算法,在保证图像质量的同时,使亚像素搜索点数减少53.0%~ 76.5%,视频序列总编码时间节省了23%~33%。     

Development of the PenTile Matrix™ color AMLCD subpixel architecture and rendering algorithms

Abstract— Color subpixel rendering is enhanced by co‐optimizing the color subpixel architecture and algorithms with respect to human vision. This has resulted in the PenTile? display technology, which provides double the information content per subpixel when compared to a conventional RGB Stripe display. Output performance results from mathematical modeling, software simulations, and prototype AMLCDs displays demonstrate significant quality improvements to both text and full‐color images in comparison to images from RGB Stripe displays that have the same number of subpixels and column drivers.     

尺寸测量中的边缘检测算法研究

主要介绍了一种改进的边缘检测算法及相应计算模板和公式，并对算法的误差进行了分析。因为传统的边缘检测算法在理论上较成熟，但在实际应用时可操作性较差。所以针对零件图像测量的实际应用情况，使用基于Sobel算子的改进的方向算子，对灰度图像进行处理得到梯度图像。综合应用其他算法，实现了对目标边缘的准确检测。然后在梯度图像上沿目标边缘的梯度方向进行多项式插值法亚像素细分计算，对目标边缘进行亚像素精确定位。在文章的最后，用实例说明了本算法的可行性。     

Super‐resolution of the undersampled and subpixel shifted image sequence by a neural network

Numerous approaches to super‐resolution (SR) of sequentially observed images (image sequence) of low resolution (LR) have been presented in the past two decades. However, neural network methods are almost ignored for solving SR problems. This is because the SR problem traditionally has been regarded as the optimization of an ill‐posed large set of linear equations. A designed neural network based on this has a large number of neurons, thereby requiring a long learning time. Also, the deduced cost function is overly complex. These defects limit applications of a neural network to an SR problem. We think that the underlying meaning of the SR problem should refer to super‐resolving an imaging system by image sequence observation, instead of merely improving the image sequence itself. SR can be regarded as a pattern mapping from LR to SR images. The parameters of the pattern mapping can be learned from the imaging process of the image sequence. This article presents a neural network for SR based on learning from the imaging process of the image sequence. In order to speed up the convergence, we employ vector mapping to train the neural network. A mapping vector is composed of some neighbor subpixels. Such a well‐trained neural network has powerful generalization ability so that it can be used directly to estimate the SR image of the other image sequences without learning again. Our simulations show the effectiveness of the proposed neural network. © 2004 Wiley Periodicals, Inc. Int J Imaging Syst Technol 14, 8–15, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.20001