Real-time CAM-based Hough transform algorithm and its performance evaluation

Abstract. This paper proposes a highly parallel Hough transform algorithm for real-time straight-line extraction and its hardware implementation on a content-addressable memory (CAM). To achieve high-speed processing, incrementation for voting, which composes the Hough transform, and calculations for coordinate updating are carried out for the every scan line, not every edge pixel, and extracting maxima in Hough space is executed by parallel comparing. Moreover, variously weighted voting achieves more accurate line extraction in spite of the quantization error and noise in the image space. In the implementation, the CAM acts as a PE (processing-element) array that effectively performs highly parallel processing for the Hough transform and also as a memory for two-dimensional Hough space, and both voting and peak extraction are directly executed by the CAM. Evaluations of CAM hardware size, processing time and the accuracy of line extraction show that a real-time and high-resolution Hough transform for a 256256 picture can be achieved using a single CAM chip with current VLSI technology. This CAM-based Hough transform algorithm promises to be an important step towards the realization of a real-time and compact image-understanding system. Received: 15 August 1998 / Accepted: 15 March 2000     

基于随机Hough变换的道路边界识别算法研究

道路边界识别是基于机器视觉的智能车辆关键技术之一,提出了一种基于直线道路模型的道路边界识别和跟踪算法,该算法包括道路边界初始识别算法和跟踪算法两部分。在初始道路边界识别阶段,对预处理后的图像进行逐行搜索道路边界候选点,并结合相关的道路约束条件,采用Hough变换对候选边界点进行拟合,以提高鲁棒性。该方法的创新点是在道路跟踪算法中,建立目标搜索区域（OSA）,OSA的宽度根据道路边界识别的拟合可信度的变换进行动态调整,并采用随机Hough变换的算法对OSA中的边界点进行拟合,这样显著减少了运算时间和存储空间。试验结果表明,该方法不仅满足道路边界识别的实时性要求,还具有较好的鲁棒性。     

Geodesic voting for the automatic extraction of tree structures. Methods and applications

This paper presents new methods to segment thin tree structures, which are, for example present in microglia extensions and cardiac or neuronal blood vessels. Many authors have used minimal cost paths, or geodesics relative to a local weighting potential P, to find a vessel pathway between two end points. We utilize a set of such geodesic paths to find a tubular tree structure by seeking minimal interaction. We introduce a new idea that we call geodesic voting or geodesic density. The approach consists of computing geodesics from a set of end points scattered in the image which flow toward a given source point. The target structure corresponds to image points with a high geodesic density. The “Geodesic density” is defined at each pixel of the image as the number of geodesics that pass over this pixel. The potential P is defined in such way that it takes low values along the tree structure, therefore geodesics will migrate toward this structure thereby yielding a high geodesic density. We further adapt these methods to segment complex tree structures in a noisy medium and apply them to segment microglia extensions from confocal microscope images as well as vessels.     

基于图像处理的晶体生长直径检测技术

针对单晶硅生长过程中的直径测量,提出一种“不在同一条直线上的三点确定圆”的直径测量算法.首先对获取到的单晶生长图像进行图像预处理,提取单晶体的边缘信息,然后基于“不在同一条直线上的三点确定圆”的原理,对部分“释热光环”提取特征点,从而获取硅单晶直径像素值,再通过量纲转化把直径像素值转化为长度值.试验表明,该算法获得的硅单晶直径的准确率满足直径检测系统的要求.     

基于统计预测的车道边缘点搜索算法

本文提出了车道连续性、直线性假设和边缘点位置的可预测性推论，实现了一种基于统计预测的车道边缘点搜索算法。该算法利用统计预测基本原理对车道边缘点的位置进行预测．将搜索限制在预测范围内，提高了搜索精度，减小了整幅图像的搜索范围，保证了算法的实时性能。     

一种异性扩散的自适应平滑滤波算法

固定尺度的高斯滤波器对图像中的所有像素使用相同的加权系数,在平滑噪声的同时也模糊了边缘信息,影响了后期的边缘定位和提取,为此提出了一种异性扩散的自适应仿高斯滤波器,通过计算当前像素与周围像素的灰度梯度,预判该像素类型（噪声点或是边缘点）,并根据像素点类型和梯度生成自适应的、各向异性的加权系数,通过多次平滑迭代实现权重的异性扩散。实验结果表明,该滤波算法具有平滑噪声的同时保留并锐化边缘效果。     

基于边缘颜色信息的车牌定位算法

车辆牌照的自动识别是智能交通系统中的一项重要技术,而车辆牌照的定位又是车牌识别的关键点之一.提出了一种基于边缘颜色点对与扫描线相结合的车牌定位方法.首先进行彩色边缘检测,然后以每一边缘点为中心,垂直于边缘方向取一条线段,在线段内检测边缘点两侧像素的颜色是否分别匹配车牌的底色与字符颜色,若是,则保留为候选车牌边缘点;通过扫描线搜索定位并分割出车牌区域.方法抓住了车牌背景与字符具有固定颜色搭配的重要特点,综合利用了车牌的结构特征和纹理特征,提高了车牌定位的可靠性.实验结果表明,算法能够实现车牌的快速精确定位.     

基于最大目标函数的医学图像边缘检测

鉴于医学图像特点和传统算法的缺点,提出一种新的医学图像边缘检测算法,该算法通过考察3×3模板的理想边缘结构特征,将模板像素沿边缘方向分为两个集合,通过距离度量函数构造适当的目标函数,计算四个方向的目标函数值,从而获得最大目标函数值,并与给定的阈值比较,应用非极大抑制方法,判别该象素是否为边缘点,最后进行双阈值连接,可得到单像素边缘图像。仿真结果表明,该算法能够有效检测医学图像边缘,所检测出的图像边缘细节丰富,单像素宽,定位准确。     

基于局部能量的边缘检测方法

本文从局部能量的角度提出了一种有效的图像边缘检测方法.在以一个像素点为中心的对称区域中,计算区域内所有像素的灰度值与中心像素的灰度值之间的差值,将差值平方的总和作为中心点所对应的局部能量.该局部能量可以有效地用于检测图像的边缘,因为边缘点的局部能量要比对应光滑区域内的像素点大得多.根据本文所构造的局部能量函数可以有效地找到边缘点.本文使用Baddeley误差度量（BEM）方法来评估本文方法检测结果的准确性.实验结果表明本文方法检测效果比较好.     

一种提取直线的随机方法

基于Hough变换提取直线的方法，由于要预先量化参数空间，因此需要很大的存储量和计算量．基于RHT(Randomized Hough Transform)提取直线的方法是通过随机选取两个点得到直线的参数，而后在参数空间对相应的参数进行累加、判断，该方法虽然无需预先量化参数空间，但是其在直线检测时，收敛速度慢．为此提出一种新的随机检测直线(Random Line Detection)的方法，在图象边缘点构成的数据空间中随机选取3个点，根据距离准则获得一条可能的直线，然后在数据空间中进一步判断直线的真实性，实验证实了该方法能有效的减少存储空间并降低计算量。     

基于计算机视觉的板类零件曲面测量系统

利用计算机视觉原理，建立了板类零件曲面测量系统，该系统首先根据人眼感知事物原理，采用神经网络来拟合图像坐标与空间坐标的映射关系；然后以光栅投影条纹为特征，用小波变换提取条纹边缘，在此基础上，提出搜索式无监督聚类方法，使带状离散边缘点按边缘实际分布情况分为不同组群，并将各组边缘点分别拟合成连续B样条曲线，同时结合视觉几何不变性，实现了亚象素级的立体精匹配；接着，运用小波分解来讲接图象，融合数据，并由图象坐标与空间坐标的映射关系，求解曲面上点的空间坐标，测量精度可控制在0．5mm/m以内。     

自适应投票快速中值滤波算法的进一步研究

以邻域像素的相似性为基础,提出了分组投票取中值法,避免了大量的零值票箱参入检票过程,提高了检票速度。还提出了投反票思想,利用窗口移动的相关性,使每次参入投票的像素个数大大减少,避免了每次投票前的票箱清理操作,优化了投票过程,提高了投票速度。消除了现有高速中值滤波算法不适合小窗口的缺陷,用较少的求和运算和比较运算,使各类窗口下的中值滤波速度大幅度提高。还可以快速获得图像的局部参数（局部极值、极值的个数、局部均值）,用这些参数作为噪声检测的依据,能够准确地判断噪声,提高中值滤波的效果。     

超声图象的自适应线边界检测方法

针对超声图象存在一种特殊的斑点噪声 ,使图象边界与细节变得模糊 ,而严重影响图象质量的问题 ,提出了一种自适应线边界检测方法 ,即以不同方向与不同大小的“窄条”来近似组织边界 ,这种“窄条”是指具有不同长度、方向各异的线段 .该方法首先将经过每一像素点的“窄条”的最大投影作为该点的灰度值 ;然后应用假设试验优化方法确定“窄条”方向 ,而“窄条”的长度由基于区域增长的局部统计特性决定 .通过引入噪声抑制与边缘保留量化参数 ,进行了仿真斑点图与组织超声图的处理实验 ,并与自适应斑点抑制滤波、直接线边界检测方法进行了比较 ,结果证明 ,该方法不但能有效地抑制斑点噪声、而且能很好地保留与增强图中的组织边界与局部细节 .     

Edge and Depth from Focus

This paper proposes a novel method to obtain the reliable edge and depth information by integrating a set of multi-focus images, i.e., a sequence of images taken by systematically varying a camera parameter focus. In previous work on depth measurement using focusing or defocusing, the accuracy depends upon the size and location of local windows where the amount of blur is measured. In contrast, no windowing is needed in our method; the blur is evaluated from the intensity change along corresponding pixels in the multi-focus images. Such a blur analysis enables us not only to detect the edge points without using spatial differentiation but also to estimate the depth with high accuracy. In addition, the analysis result is stable because the proposed method involves integral computations such as summation and least-square model fitting. This paper first discusses the fundamental properties of multi-focus images based on a step edge model. Then, two algorithms are presented: edge detection using an accumulated defocus image which represents the spatial distribution of blur, and depth estimation using a spatio-focal image which represents the intensity distribution along focus axis. The experimental results demonstrate that the highly precise measurement has been achieved: 0.5 pixel position fluctuation in edge detection and 0.2% error at 2.4 m in depth estimation.     

基于边缘和局部熵融合的视觉安全性评估方法

针对基于局部熵进行加密图像视觉安全性评估存在块效应的局限性，引入图像的边缘特征，通过共有边缘来衡量加密图像与原始图像的边缘相似度，消除了块效应。由于局部熵对加密等级低的图像不敏感，边缘相似度对加密等级高的图像不敏感，将两个评估方法进行自适应融合，提出[SLEES]（Local Entropy and Edge Similarity，[SLEES]）指标。通过改变图像像素位置和图像像素值的加密方式处理图像和视频帧进行测试，实验结果表明，[SLEES]指标相比传统评估指标有更好的鲁棒性，评估范围更广。     

基于曲率半径的数据分割

数据分割是反向工程几何建模的关键一步,该文在对散乱点建立正交模型的基础上,利用参数曲面拟合法求解数据点的局部曲率,进而求出各点的局部平均曲率半径,根据曲率半径的阀值可以检测出各类边点,包括楞边、光滑过渡的峰线（ridge lines）和谷线（valley lines）,然后进行边点连接形成封闭的边环,最后利用扫描线法把整体数据分为多个相互独立的子集。     

基于FPGA的拼接式CMOS线阵相机系统设计

针对单个线阵相机在大视场下传感器边缘失光的问题,提出了在保证分辨率的前提下使用多个相对较低像素的相机拼接,对视场进行分担的方法;设计了一种基于FPGA的拼接式CMOS线阵相机系统,系统整体像素为4 096Pixels,行频为2.5kHz;详细介绍了系统的工作原理、硬件构成和各芯片的程序设计,实现了基于FPGA的多相机图像实时校正算法,并将校正后的图像通过高速USB芯片传输到采集软件;实验结果表明,在视场较大的条件下同等像素的拼接式相机相比单相机在视场边缘具有更好的成像质量,且需要的物距更短,可应用于对多离散目标的检测。     

A Bayesian approach to the Hough transform for line detection

This paper explains how to associate a rigorous probability value to the main straight line features extracted from a digital image. A Bayesian approach to the Hough Transform (HT) is considered. Under general conditions, it is shown that a probability measure is associated to each line extracted from the HT. The proposed method increments the HT accumulator in a probabilistic way: first calculating the uncertainty of each edge point in the image and then using a Bayesian probabilistic scheme for fusing the probability of each edge point and calculating the line feature probability.     

Pixel and sub-pixel accuracy in geometrical correction of AVHRR imagery

Abstract

This paper makes two contributions to the problem of geometrical rectification of Advanced Very High Resolution Radiometer (AVHRR) image data. The first is an automatic procedure for the location of ground control points (GCPs) without the involvement of operator intervention and using a library of 'chips’ of AVHRR data for the GCPs. The second involves the adaptation of a method previously used by Torlegard for Landsat Multi-spectral Scanner (MSS) data to the case of AVHRR data to attempt to achieve sub-pixel accuracy in the rectification; we have achieved an accuracy of 26 per cent of the pixel edge and have every hope that this can be reduced to below 20 per cent in the future.