多摄像机图像拼接视觉归一化技术研究

视觉归一化是多视点图像拼接领域的一个关键技术,在对大量图像处理算法研究的基础上,提出了一种针对多摄像机图像拼接的视觉归一化处理方法。该方法主要包括图像颜色校正和图像边缘融合两个模块;在图像颜色校正模块中,引入了图像区域划分策略和自适应颜色调节因子,使不同的像素点都有不同的颜色调节因子,并充分利用相邻图像间的颜色关联性对目标图像的颜色进行自适应校正;在图像边缘融合模块中,利用反映射矩阵计算出拼接图像的重叠区域,利用自适应边缘融合因子对重叠区域进行边缘融合处理。实验结果表明,该方法能够较好地减少甚至消除拼接图像间的视觉差异,较好地改善了图像拼接的视觉效果。     

面向光流分析的机器人红外视觉图像对预处理

为了增强机器人单目视觉系统在黑暗环境下的环境感知能力,提高其在视觉图像序列处理过程中光流场计算的准确率,对由红外摄像机所采集的红外图像序列的预处理方法进行了研究,建立了一种空域与变换域相结合的处理方法。为平衡红外图像序列对之间的亮度差异并提高图像亮度,在空域中采用直方图均衡化方法进行处理;然后,对图像进行非下采样Contourlet变换,在变换域中利用图像的强边缘、弱边缘和噪声在不同分解尺度和分解方向上具有不同几何流的性质进行区分,分别采取保留、增强以及去除处理;最后利用调整后的系数进行图像重构。该预处理方法在平衡图像亮度的同时增强了图像纹理并减少了噪声。实验结果表明,预处理程序有效地提高了红外图像对的光流有效点识别率,该方法增强了机器人单目视觉系统在黑暗环境中对环境的感知能力。     

像素不扩展视觉密码的边缘增强研究

针对像素不扩展视觉密码的边缘恢复失真问题,在分析图像边缘特征的基础上,通过构造LP算子,设计了一种具有可变膨胀倍数的边缘增强算法,并给出了边缘增强的像素不扩展视觉密码方案的设计方法。实验结果表明,该方法有效地改善了边缘的恢复质量,并显著提高了整幅图像的视觉效果。     

大规模立体显示墙系统的构建

从硬件选择、几何校正、边缘融合、颜色校正和软件环境等5方面系统地研究了构建大规模立体显示墙系统的关键技术.采用并行绘制技术和基于计算机视觉的自动校正,由Linux集群驱动多台普通投影仪拼接成一个统一的高分辨率、大规模的被动立体显示系统.     

基于双变换的红外与可见光图像融合增强

通过研究可见光视觉图像与红外图像序列的相关特性及融合处理中存在的问题, 提出一种基于视觉图像预增强处理的曲波与小波双变换融合机制来有效保留边缘信息和图像细节信息。实验结果表明所提出的方法提高了融合图像的视觉感知质量, 为遮挡和伪装目标的检测和定位提供了更为有效的空间环境, 并且融合增强后的图像具有更高的信噪比和更少的人工痕迹, 与当前最新的其他方法相比是一种具有更好融合质量和更高效率的图像/视频融合方法。     

基于眼动跟踪的视频视觉感兴趣区域提取仿真

海量视频中具有显著特征的区域为感兴趣区域。由于视频图像中普遍存在较多单目标特征,轮廓边缘关联性较强,干扰视觉区域的提取。于是提出基于眼动跟踪的视频视觉感兴趣区域提取方法。提前转换原始视频颜色空间,降低色彩影响程度,检测原始视频边缘,利用形态学闭运算填补视频边缘断裂位置并去噪,通过旋转校正处理得到简单完整的视频图像,利用眼动跟踪处理技术提取纹理特征以及亮度显著图,实现视频视觉感兴趣区域提取。实验结果表明,研究方法能够精准提取出视频视觉感兴趣区域,峰值信噪比最高可达1.5dB,验证了该方法的应用精度和图像质量增益均较高。     

改进的灰度校正算法在路面裂缝图像预处理中应用

针对传统的灰度校正算法在对亮度不均匀的路面裂缝图像处理时容易造成边缘模糊,局部对比度得不到增强的问题,本文在传统的快速灰度校正算法基础上,提出了一种改变初始背景图像选取方式以及采取局部对比度增强图像和整体灰度校正图像相互融合的改进型灰度校正算法.实验结果表明,与传统方法相比,改进后的算法对路面裂缝类图像的预处理效果明显增强.     

基于生物视觉特征的目标轮廓提取算法

自然场景中的目标轮廓提取是计算机视觉中的一个重要研究问题。其难点在于场景中大量的纹理边缘严重地干扰了轮廓提取的完整性。近年来,一些研究工作将生物视觉特征引入图像边缘轮廓提取,取得了一定的效果。其中通过引入视觉外区抑制特征可以在提取物体轮廓边缘的同时抑制一定量的纹理边缘,从而得到轮廓边缘集合。然而在整合轮廓边缘时,传统模型仅仅采用求交并集的简单合并方法,使得强响应的细小纹理残留。基于此,提出了一种改进的基于生物视觉特征的自然场景目标轮廓提取算法。首先采用多水平抑制方法得到候选轮廓边缘集合。接着将一种基于生物视觉特征的边缘组合方法用于将候选边缘整合成为一个完整的目标轮廓。与传统的外区抑制算法相比,基于视觉特征的轮廓提取算法提高了自然场景中目标轮廓提取的准确性和完整性。     

一种改进 Hough 变换的车道线检测算法*

对基于单目视觉的车道线检测技术进行研究,针对复杂环境中车道线检测易受图像中其它直线段的影响而出现较多误检测的问题,提出一种改进算法。首先将车道线图像灰度化,采用中值滤波去除图像噪声;再次,采用Canny算子和OTSU方法进行边缘增强和图像二值化;最后,使用改进的Hough变换进行车道线检测。实验结果表明,该方法可以有效地滤除其它直线段的影响,在复杂环境中获得良好的车道线检测效果。     

视觉位置控制系统的线性化校正

本文针对视觉定位控制问题，讨论了机器视觉和受控装置的非线性影响，分析了视觉反馈控制系统中各环节的信息变换关系，用分块线性化方法分析了视觉图像的校正原理和线性回归算法对视觉控制校正的可行性，介绍了实施校正所采取的关键技术及实验方法。     

A self-organization learning algorithm for visuo-motor coordination in unstructured environments

In this paper, we propose a learning algorithm for coordinating a robot system where the movement of an arm is controlled through a stereo camera system. Instead of calibrating the usually complex non-linear transformation between the arm and cameras, the algorithm decomposes the whole transformation automatically into local linear transformations and then makes the linearization map recorded by the arm controller. The linearization is carried out by a learning process based on a Kohonen-style self-organization network. To deal with unstructured environments in which some obstacles exist, some virtual forces are introduced for dealing with the high degree of complexity underlying in the transformation. This work was presented, in part, at the International Symposium on Artificial Life and Robotics, Oita, Japan, February 18–20, 1996     

Portable autonomous walk calibration for 4-legged robots

In the present paper we describe an efficient and portable optimization method for calibrating the walk parameters of a quadruped robot, and its contribution for the robot control and localization. The locomotion of a legged robot presents not only the problem of maximizing the speed, but also the problem of obtaining a precise speed response, and achieving an acceptable odometry information. In this study we use a simulated annealing algorithm for calibrating different parametric sets for different speed ranges, with the goal of avoiding discontinuities. The results are applied to the robot AIBO in the RoboCup domain. Moreover, we outline the relevance of calibration to the control, showing the improvement obtained in odometry and, as a consequence, in robot localization.     

Calibration of a magnetometer array using motion capture equipment

This paper studies the problem of calibrating an array of single‐axis magnetometers in an unknown static inhomogeneous magnetic field using motion capture equipment. A proof of identifiability is given, practical identifiability of calibration parameters is established in simulation, and real world experiments are conducted to demonstrate the feasibility of this approach. Unlike many state of the art techniques, the proposed solution does not require a homogeneous field, as in fact, we demonstrate that an inhomogeneous field enlarges the set of identifiable parameters. Under the above‐mentioned assumptions, this approach may be used to extend self‐calibration techniques of visual‐inertial setups to magnetic sensor arrays in indoor environments.     

Feature extraction for outdoor mobile robot navigation based on a modified Gauss–Newton optimization approach

This paper discusses the problem of feature detection for semi-structured outdoor environments such as campuses and parks using laser range sensors. In these environments, commonly encountered natural features that can be very useful for mobile robot navigation include edges (large discontinuity) and circles (e.g., trees, pillars). The term feature is used to denote objects which are “likely” to be detectable when the sensor is moved to new locations. Note that there has been no systematic approach for feature detection in outdoor environments. In this paper, we present an algorithm for feature detection. The algorithm consists of data segmentation and parameter acquisition. A modified Gauss–Newton method is proposed for fitting circle parameters iteratively. Experimental results show that the proposed algorithm is efficient in detecting features for semi-structured outdoor environments and is applicable to real time simultaneous localization and mapping.     

Hierarchical stochastic production planning for flexible automation workshops

This paper presents the hierarchical stochastic production planning (HSPP) problem for flexible automation workshops (FAWs) in agile manufacturing environments, which is a multiple-period multiple-product problem with random material supply, demands, capacities, processing times, rework and waste products. To solve the HSPP problem, a mathematical model is built up first. Then, an algorithm for HSPP is deduced in detail by using a stochastic interaction/prediction approach. The corresponding software package named as stochastic interaction/prediction algorithm (SIPA) has been developed and is presented in this paper, through which examples of HSPP have been studied, and which show that the algorithm can optimally decompose medium-term random product demand plans of an FAW into short-term stochastic production plans to be executed by FMSs in the FAW. Finally, the application of the algorithm is presented in detail through one of those examples.     

求解对称带状广义特征值问题的扩展分治算法

提出了分布式环境下计算对称带状广义特征值问题的一种扩展分治算法，给出了特征值分割定理及其证明．算法在扩展分治的基础上，利用二分压缩结合广义Rayleigh商迭代计算广义特征对．理论分析和数值实验表明，对于窄带宽大规模的广义特征值问题，该分治算法明显优于LAPACK软件包．结合并行性好的多分法，在分布式环境下获得了很好的并行效果．     

非平稳环境下基于小波变换的图像去噪

传统的图像去噪算法往往仅对平稳或缓慢变化的噪声有效,且残留的图像噪声较大。对此,研究了非平稳环境下基于小波变换的图像去噪算法。该算法根据图像与噪声在小波域的分布特性以及它们的小波变换模极大值随尺度的变化大小不同,运用迭代算法得到不同尺度小波域中噪声的具体位置以及小波系数大小,完成了图像去噪。实验结果表明,对峰值信噪比较低的图像,该方法去噪后峰值信噪比比传统方法的高,并且保留了较多的图像细节,同时对平稳和非平稳的噪声都能进行较好地去噪。     

基于直线约束的非量测数码相机标定

提出一种非量测数码相机标定的直接方法。该方法基于直线约束条件,即在正确进行畸变纠正后,物方空间的三维直线投影到像片平面上也应该是一条直线。为了将畸变直线纠正成理想直线,利用畸变模型对其进行畸变纠正,并采用Levenberg-Marquardt算法对非线性方程组进行求解,解出符合条件的最佳畸变参数。实验结果证明该方法具有较强的鲁棒性与实用性。     

Online Learning and Adaptation of Autonomous Mobile Robots for Sustainable Agriculture

In this paper we will introduce the application of our newly patented double hierarchical Fuzzy-Genetic system (British patent 99-10539.7) to produce an intelligent autonomous outdoor agricultural mobile robot capable of learning and calibrating its controller online in a short time interval and implementing a life long learning strategy. The online and life long learning strategy allow the outdoor robots to increase their experience and adapt their controllers in the face of the changing and dynamic unstructured outdoor agricultural environments. Such characteristics permit prolonged periods of operation within dynamic agricultural environments, which is an essential feature for the realization of a platform vehicle for use in sustainable agriculture and organic farming.     

基于灭点的倾斜畸变车牌图像快速校正方法研究*

由于相机与汽车牌照之间的方位关系导致了拍摄图像发生倾斜畸变,极大影响了车牌识别率。针对这一问题,利用畸变图像所蕴涵的灭点信息,提出了一种快速校正方法：首先对图像进行预处理,提取车牌的四条边框来产生灭点;然后根据灭点信息快速标定相机内参数并计算空间平面重建矩阵;最后进行空间坐标变换和灰度插值完成畸变图像的校正。实验表明,该方法对于倾斜畸变车牌有良好的校正效果,并有效提高了识别率。