Skeleton body pose tracking from efficient three-dimensional motion estimation and volumetric reconstruction

We address the problem of body pose tracking in a scenario of multiple camera setup with the aim of recovering body motion robustly and accurately. The tracking is performed on three-dimensional (3D) space using 3D data, including colored volume and 3D optical flow, which are reconstructed at each time step. We introduce strategies to compute multiple camera-based 3D optical flow and have attained efficient and robust 3D motion estimation. Body pose estimation starts with a prediction using 3D optical flow and then is changed to a lower-dimensional global optimization problem. Our method utilizes a voxel subject-specific body model, exploits multiple 3D image cues, and incorporates physical constraints into a stochastic particle-based search initialized from the deterministic prediction and stochastic sampling. It leads to a robust 3D pose tracker. Experiments on publicly available sequences show the robustness and accuracy of our approach.     

基于视觉的驾驶机器人导航技术

对驾驶机器人的视觉导航技术进行了研究,提出了一种快速可靠的车辆引导线检测算法.该算法首先对感兴趣的移动窗口进行有条件的边缘检测,进而引入了梯度方向直方图的处理思想,制定了车辆引导线边缘候选点集与有效点集的筛选算法,最后,通过所提出的一种基于标量化处理Kalman滤波方法实现了引导线特征参数的快速准确提取.试验表明,本文提出的检测算法能够满足复杂路况下的导航需要,具有30ms/帧的实时处理速度.     

An algorithm for finger-vein segmentation based on modified repeated line tracking

Abstract

Image segmentation is an important step for finger-vein identification technique. However, it is difficult to extract precise details of the image because of the irregular noise and shades around the finger-vein. The repeated line tracking algorithm achieves good segmentation performance for low quality images of finger-vein, but it has some drawbacks such as low robustness and efficiency. In this paper, a modified repeated line tracking algorithm is proposed for image segmentation of finger-vein. Firstly, we propose a segmentation method called threshold image to execute rough segmentation and obtain binary and skeleton image of finger-vein. Secondly, the width of finger-vein is estimated based on the binary and skeleton image. The parameters are revised according to the width. Then, the modified repeated line tracking algorithm is executed to figure out the locus space of finger-vein based on the revised parameters. Finally, processing results are obtained by using Otsu algorithm which executes exact segmentation on the locus space. Experiments show that the proposed algorithm is more robust and efficient than traditional repeated line tracking algorithm.     

Pose estimation from a two-dimensional view by use of composite correlation filters and neural networks

We present a technique to estimate the pose of a three-dimensional object from a two-dimensional view. We first compute the correlation between the unknown image and several synthetic-discriminant-function filters constructed with known views of the object. We consider both linear and nonlinear correlations. The filters are constructed in such a way that the obtained correlation values depend on the pose parameters. We show that this dependence is not perfectly linear, in particular for nonlinear correlation. Therefore we use a two-layer neural network to retrieve the pose parameters from the correlation values. We demonstrate the technique by simultaneously estimating the in-plane and out-of-plane orientations of an airplane within an 8-deg portion. We show that a nonlinear correlation is necessary to identify the object and also to estimate its pose. On the other hand, linear correlation is more accurate and more robust. A combination of linear and nonlinear correlations gives the best results.     

Improving thin-film manufacturing yield with robust optimization

A novel robust optimization algorithm is demonstrated that is largely deterministic, and yet it attempts to account for statistical variations in coating. Through Monte Carlo simulations of manufacturing, we compare the performance of a proof-of-concept antireflection (AR) coating designed with our robust optimization to that of a conventionally optimized AR coating. We find that the robust algorithm produces an AR coating with a significantly improved yield.     

Pose estimation for autonomous grasping with a robotic arm system

A robust and accurate method for estimating the 3-D pose of a planar rigid object is presented. This article demonstrates that 3-D pose estimation becomes feasible by using the 2-D tracking points on an object of scale-invariant feature transform (SIFT) and 3-D point cloud detected by stereo vision on an object, assuming that a 3-D geometric model of an object is known a priori. The roll and pitch angles of an object are estimated by the normal vector of approximate plane of 3-D point cloud on an object and the yaw angle is estimated by 2-D tracking point on an object of SIFT. Accurate object detection and localization in the camera coordinate system is crucial for grasping. In the motion planning, the bidirectional rapidly exploring random tree algorithm is used to search for a valid path for efficient grasping. Our robot arm can robustly and autonomously grasp a randomly rotative rigid object detected by SIFT in 3-D space. We have realized a grasping scenario with a dexterous arm (ADAM) such that an object in front of ADAM can be grasped. This demonstration shows how the proposed components build a dexterous and robust system integrating object detection, pose estimation, and motion planning.     

Large production line optimization using simulated annealing

We present a robust generalized queuing network algorithm as an evaluative procedure for optimizing production line configurations using simulated annealing. We compare the results obtained with our algorithm to those of other studies and find some interesting similarities but also striking differences between them in the allocation of buffers, numbers of servers, and their service rates. While context dependent, these patterns of allocation are one of the most important insights which emerge in solving very long production lines. The patterns, however, are often counter-intuitive, which underscores the difficulty of the problem we address. The most interesting feature of our optimization procedure is its bounded execution time, which makes it viable for optimizing very long production line configurations. Based on the bounded execution time property, we have optimized configurations of up to 60 stations with 120 buffers and servers in less than five hours of CPU time.     

基于HT的显微视觉亚像素定位算法

根据显微图像的几何特点,提出了基于HT（Hough Transform）十字显微图像亚像素定位算法（SLACHT）和基于HT近似圆显微图像亚像素定位算法（SLARHT）,对显微图像特征中的直线和近似圆的亚像素定位做了深入的研究。通过与经典直线HT定位算法和经典随机HT圆定位算法进行实验对比,验证了该定位算法的优越性;通过与Matlab中的直线HT定位算法和圆HT定位算法进行实验对比,验证了该算法的可靠性。大量实验表明,该算法的定位精度优于0．01像素,具有运行速度快、鲁棒性强的特点。     

Iterative algorithm for optimal fiducials under weak perspective projection

In previous work, we designed space fiducials with the aim of making camera pose determination as noise‐insensitive as possible. These fiducials turned out to be sets of points that formed concentric regular polyhedra. Here, we apply an idea of Dementhon and Davis and test and analyze an iterative linear algorithm in conjunction with our optimal fiducials to increase the accuracy of the computed camera pose. We also analyze under what circumstances this iterative algorithm is guaranteed to converge to the correct solution. Comprehensive computer simulations illustrate the behavior of the algorithm and the degree of improvement in pose determination in case of convergence. © 2009 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 19, 27–36, 2009     

A robust and computationally efficient algorithm for mean scattererspacing estimation

Mean scatterer spacing (MSS) has been recognized to be a useful tool for tissue characterization. Most of the work in this area either uses the amplitude or the phase information of the spectrum of the backscattered ultrasound echo to estimate the MSS. Simulations have shown that the latter approach is more robust in the presence of irregularities in the scatterer distribution. However, most of the algorithms based on the phase information of the spectrum are computationally demanding and cannot be used in real-time. We present a computationally efficient and robust algorithm which uses the magnitude and phase information of the spectrum to estimate the MSS. This algorithm exploits the spectral redundancy present in the backscattered echo signal by generating spectral lines through a nonlinear (quadratic) transformation of the RF echo signal. Results of simulations comparing the performance of the proposed algorithm and previous approaches from the literature are presented to demonstrate the robustness of the proposed algorithm. Experiments involving phantoms and in vitro tissue samples are also presented. The feasibility of implementing a real-time MSS imaging system based on the proposed method is discussed     

一种改进NMF算法及其在人脸识别中的应用

为了提高非负矩阵分解(NMF)算法对光照、姿态等外部因素的鲁棒性,本文对传统的NMF进行改进,提出了一种改进的NMF方法.首先对NMF基图像进行判别分析,然后选择主要反应类内差异的基图像来构造子空间,最后在子空间上进行识别.通过Havard人脸库和Umist人脸库上的实验,结果表明,该方法能够对光照和姿态的变化具有一定的鲁棒性和较高的识别率,比传统的NMF方法和PCA等子空间分析法识别率提高了20%以上.     

Image compression by vector quantization with noniterative derivation of a codebook: applications to video and confocal images

We demonstrate an image-compression technique that uses what we believe is a new noniterative codebook generation algorithm for vector quantization. The technique supports rapid decompression and is equally applicable to individual images or to a set of images without the need for interframe processing. Compression with a single-image codebook is tested on (1) ten confocal images of the hindbrain of a mouse embryo, (2) video images of a polystyrene microsphere that is manipulated by a focused laser light, and (3) five fluorescence images of the embryo eye lens taken at different magnifications. The reconstructions are assessed with the normalized mean-squared error and with Linfoot's criteria of fidelity, structural content, and correlation quality. Experimental results with single-image compression show that the technique produces fewer local artifacts than JPEG compression, especially with noisy images. Results with video and confocal image series indicate that single-image codebook generation is sufficient at practical compression ratios for producing acceptable reconstructions for mouse embryo analysis and for viewing optically trapped microspheres. Experiments with the magnified images also reveal that the compression scheme is robust to scaling.     

Multilevel quality-guided phase unwrapping algorithm for real-time three-dimensional shape reconstruction

A multilevel quality-guided phase unwrapping algorithm for real-time 3D shape measurement is presented. The quality map is generated from the gradient of the phase map. Multilevel thresholds are used to unwrap the phase level by level. Within the data points in each level, a fast scan-line algorithm is employed. The processing time of this algorithm is approximately 18.3 ms for an image size of 640x480 pixels in an ordinary computer. We demonstrate that this algorithm can be implemented into our real-time 3D shape measurement system for real-time 3D reconstruction. Experiments show that this algorithm improves the previous scan-line phase unwrapping algorithm significantly although it reduces its processing speed slightly.     

一种用于风机故障诊断的免疫克隆特征选择算法

提出一种新的用于风机故障诊断的免疫克隆特征选择算法.提取了生产线上实测风机噪声的时域波形结构特征、小波分析特征及听觉谱特征,进行特征选择和故障诊断仿真实验.实验结果表明：在特征选择后的特征数目比原特征数目减少61% 的情况下,支持向量机分类器的分类正确率下降很小,分类时间显著减少.实验结果证明了该算法的有效性和鲁棒性,且能有效地应用于风机故障诊断.     

A new method for the placement of buffers in serial production lines

The placement of buffers in a production line is an old and well-studied problem in industrial engineering/operations research that is still relevant today. Decisions regarding the amount and placement of buffers in a production line occur repeatedly in system design. In this paper we document a new buffer placement method for serial production lines that operate under a variety of assumptions. The method uses information generated in a production line simulation, whose conceptual representation of job flow and workstation interaction can be described with a network, to place buffers in order to maximise throughput. This buffer placement method is very efficient and can be implemented in a spreadsheet. We demonstrate the effectiveness of the method by comparing our results against those produced by a genetic algorithm for buffer placement. Experiments are conducted on a variety of test cases. This new buffer placement optimisation method will permit designers to quickly and effectively evaluate many design alternatives and thus improve final production system performance.     

Adaptive split spectrum processing using a neural network

The polarity thresholding algorithm for split spectrum processing (SSP) is known to work well once properly tuned. However, there are several problems related to the finding of the right split parameters such as the number of filters and the information carrying spectral range. Here we show that the polarity thresholding method can be formulated as a multilayer perceptron (MLP) neural network with binary neurons and binary input signals operating in feedforward mode. Then the method is generalized to process nonbinary data using an adaptive MLP with graded neurons. Experiments with real ultrasonic NDE signals are presented using the conventional backpropagation optimization algorithm (BP) and a second order optimization method (BFGS) with exact line search. Finally, alternative adaptive algorithms based on a decomposition of the network into single neurons or linear discriminants are briefly discussed.     

一种高效的实时视觉姿态测量

由光学传感器和计算机视觉理论来确定目标物体的姿态参数具有重要的实用价值和理论意义.本文以提高当前物体姿态求解方法的实时性和准确性为目的,提出一种高效的基于特征点的姿态求解方法,该方法首先采用P3P算法求解目标物体的近似姿态,然后将该近似姿态作为初始值嵌入正交迭代(OI)优化算法求取目标姿态的最优值,该方法可以适用于任意...     

一种基于模式识别的新型双端检测器

本文提出了一种基于最近邻线分类器的新的双端检测器（DTD）。主要的思想是充分地利用特征信息以及用模式识别方法来设计DTD。本文从模式分类的角度分析了二种主要的传统DTD（Geigel和相关DTD）并给出了新的设计方法。一种称为NNL分类器的新的非参数分类器被用来检测双端通话。NNL分类器具有低运算量和优良的性能。用NNL分类器．我们熔合了几种传统的DTD并且避免了存在于大多数传统DTD中的固定阈值带来的问题。因此NNL-DTD在各种条件下是鲁棒的。实验结果也显示出了这个方法比传统方法更有效。     

一种高效的弱目标线谱检测算法

为实现弱目标线谱检测,在自适应线谱增强(Adaptive Line Enhancement,ALE)算法的基础上,结合频域批处理技术,提出了一种能降低计算量的高效线谱检测算法——归一化频域批处理最小均方(Normalized Frequency-domain Block Least Mean Square,NFBLMS)算法;所提NFBLMS算法在权值迭代过程中,步长参数不受输入信号功率的影响。理论分析和数值仿真结果表明:相比于已有的线谱检测算法,NFBLMS算法能较好地解决ALE算法实时处理运算量问题,并可获得较高的系统增益,且其步长参数具有较强的鲁棒性,能同时兼顾算法的收敛速度和稳态误差。因此NFBLMS算法更适合实际工程应用。     

一种鲁棒的线纹尺图像刻线提取算法研究

为解决光照不均匀、锈斑、污损等因素造成的钢卷尺图像检测识别不准确问题,提出了一种基于Cabor变换的钢刻线提取方法。通过Cabor变换,使特征方向与特定频率的线条结构被过滤呈现出来,基于这些线条结构计算图像中的线纹尺待检刻线位置,之后通过RANSAC算法拟合刻线的双边缘,从而准确获得刻线中心位置。经实验证明,该算法可以在刻线质量较差的情况下,鲁棒地识别破损和有锈斑的刻线,有效地提高检测准确度和整个测量系统的鲁棒性,具有技术参考价值。