边缘检测算子在断口图像识别中的应用

Application of Edge Detection Operators in Recognition and Classification of Fracture Surface Images

赵莉萍 黎明

边缘检测技术是图像处理和计算机视觉等领域最基本的技术。不同的微分算子对边缘的敏感程度是不同的，如何选用适当的边缘检测算子得到准确的边缘信息，就成为断口图像处理的关键步骤。通过MATLAB进行仿真研究，比较和分析几个不同算子的边缘处理结果，从中得出Canny算子较适用于断口图像的识别与分类过程。[著者文摘]

The technique of the image edge detection is the essential technique in the field of image process. Sensitivity is different for different differential operators ,how to choose a suitable edge detection operator to obtain exact information of edge.is the essential step for fracture surface image. Be simulated by MATLAB,the functions of the different operators＇ processing are compared and analyzed ,Canny operator is found quite suitable for recognition and classification of fracture surface images.[著者文摘]

一种改进的Canny算子边缘检测算法

An improved Canny edge detection algorithm

周晓明 马秋禾 肖蓉 杨娜

介绍Canny算子边缘检测的基本原理，并对其性能进行分析和评价。针对传统Canny算子在滤波过程中存在的缺陷，提出一种基于自适应平滑滤波的改进Canny边缘检测算子，该算法根据图像中像元灰度值的突变特性，自适应的改变滤波器的权值，在平滑图像的过程中使图像的边缘锐化。在计算梯度幅值的时候采用了邻域的梯度幅值计算方法，考虑了像素对角线方向的梯度计算，进一步抑制了噪声的影响。通过对实验图像的分析表明，改进的检测算法对图像边缘提取具有较好的检测精度和准确性。[著者文摘]

This paper introduces fundamental theory of Canny operator, and carrys on its analysis and evaluation. On this foundation, an improved Canny edge detection algorithm based on adaptive smooth filtering is proposed. According to characteristic of revulsion every point gray scale, this algorithm adaptively changescoefficients of the filter, according to gradient on each point. The results of the experiment pictures indicate that the improved algorithm has the better accuracy and precision in the edge detection.[著者文摘]

基于圆邻域和环算子的抗噪边缘检测方法

Anti-noises edge detection by using circular neighborhood and ring operator

谢松法[1] 彭嘉雄[2] 何南忠[1] 施保昌[1]

针对噪声图像的边缘检测问题，提出了一种基于圆邻域和环算子的边缘检测方法．首先，在像素点的一个适当的圆邻域内取若干个同心圆环，将这些圆环上的灰度值按径向进行加权平均，得到该像素点周围一圈各等距点上的平均灰度值，这些灰度值可构成一个周期序列；然后根据Sobel和Prewitt等算子的设计思想，构造出一类高通的环算子，利用离散傅里叶变换计算周期序列与环算子的循环卷积，其最大值则作为像素点是否为边缘点的判据．实验结果表明该算法对于含有噪声的图像具有较强的鲁棒性．[著者文摘]

An edge detection approach to noisy images is proposed by circular neighborhood and ring operator. Certain concentric circles were taken in a suitable circular neighborhood of each pixel point and the weighted average intensities of the point, which could be regarded as a periodic sequence, were calculated. According to the idea of Sobel and Prewitt operator, a class of ring operator was designed. Finally, the circulation convolution of the periodic sequence and the ring operator was calculated by the discrete Fourier transform （DFT）, and the ＂gradient＂ of each pixel was obtained according to its maximum value. The approach was carried out on synthetic and nature images corrupted by Gaussian noise. The results show that the algorithm is better than standard Sobel gradient operator and LoG algorithm in terms of visual appearance of edges. The algorithm is suitable for a wide range of images and is robust in a variety of noisy situations.[著者文摘]

基于高斯Laplace算子图像边缘检测的改进

The Improvement of Image Edge Detection Based on Gauss_Laplace Operator

黄剑玲[1] 邹辉[2]

在分析高斯Laplace算子边缘检测原理的基础上，通过设置合理的模板参数，提出了一种改进的高斯Laplace模板，该模板设置的权值具有各向同性、中心点值设置为整数、不同类型方向上设置的权重不等而相同类型方向上设置等权重、并且整个模板无零值点等特点。通过VC＋＋编程使用不同的模板对同一幅图像进行实验．实验结果表明改进的模板能较准确地检测出图像边缘，检测效果优于其他模板。[著者文摘]

Based on the analysis of Gauss_Laplace operator edge detection principle, this article advances a upswing Gauss_Laplace templet from setting reasonable templet parameter. The value is same character in all orientation, central value is integer, the value is vary in different type orientation , the value is equality in same type orientation and there is not zero value in the templet, etc. We made a experiment ,which was checking the edge of a picture with different templets by VC＋＋ programme, and the result showed that the upswing one could detect it more truly .And its effect is better than others.[著者文摘]

边缘检测算子在汽车牌照检测中的研究

Study of Edge Detection Operators in Detection of Car License Plate

黄剑玲[1] 邹辉[2]

图像边缘检测一直以来都是图像处理与分析领域的研究热点。本文对数字图像处理中几种具有代表性的边缘检测算子进行了理论分析，并通过VC＋＋编程用各种算子对车牌图像进行边缘提取，通过比较得出了各种算子的精度及在检测中算子选定的参考条件，为实际应用采用哪种算法提供对照和参考。[著者文摘]

Image edge detection is always study focus in the field of image processing and analysis. This paper analyzed several typical methods of edge detection in digital image processing, and used all kinds of operators to detect the edge of a car license plate in VC＋＋. Through comparing we worked out the precision of each operator and in which time a operator is equal to. This paper provided the comparison and the reference for using which algorithm to the practical application.[著者文摘]

一种基于Sobel分解算子的图像边缘检测并行算法

A Parallel Algorithm of Image Edge Extraction Based on the Sobel Decomposing Model

付光远

串行Sobel梯度算子边缘检测算法需要将两个掩模S1和S2分别在图像的每个像素上移动．并在每个像素上进行11次加法运算，即需要11xN^2次加法，时间复杂度为O（N^2）；文章提出了一种Sobel算子分解模型。并设计了一种在SIMD—MPP模型上基于Sobel分解算子的并行图像边缘检测算法．该并行算法总共只需要8次平移操作和9次加法运算即可完成，其时间复杂度为O（1），加速比达到N^2，大大地提高了基于Sobel算子的图像边缘特征提取的效率。[著者文摘]

The image edge extraction on the basis of sequence Sobel gradient operator must move two mask S1 and S2 on every pixel of image and add 11 times, i.e., the Sobel operator requires addition 1xN^2 times. The time complexity is O （N^2）. In this paper, a decomposing model of Sobel operator is proposed, and a parallel algorithm of image edge extraction for Sobel decomposing model based on SIMD-MPP model is designed. The parallel algorithm only requires data movement 8 times and addition 9 times. The Time complexity is O（1）. The speedup of the parallel algorithm is N2. Thus, the efficiency of image edge extraction based on Sobel operator promote greatly.[著者文摘]

基于FPGA的图像边缘检测

肖健华 王思明

简要介绍了Sobel算子，给出了基于FPGA的图像边缘检测器的设计方案，分析了选用FPGA器件的优点，最后给出了系统中滤波器的设计结构和比较器、边界判断器的软件设计代码。[著者文摘]

基于敏感区域多结构元素形态学边缘检测算法

Edge Detection Algorithm for Sensitive Region Based on Morphology of Multi-Structural Element

伯绍波 闫茂德 贺昱曜

在边缘检测算法的应用中，很难同时兼顾图像处理效果和处理速度，为此，提出了一种新的边缘检测算法，将边缘处理集中在感兴趣的图像区域中。该算法利用梯度算子获得图像中的敏感区域，再构造多种结构元素，结合形态学梯度和OTSU分割法检测敏感区域的边缘。应用于沥青路面裂缝图像检测，实验结果表明，与其它边缘检测算法相比，该算法不仅具有很好的边缘提取能力，而且具有很强的抗噪能力。在保证处理效果的同时，也保证了处理速度，有很高的实用性和推广性。[著者文摘]

In the practical application of edge detection, it is difficult to make a balance between image processing effect and processing speed. In order to solve this problem, a novel edge detection algorithm is presented which the interested regions of the image are processed only. The proposed algorithm employs the gradient operator to gain the sensitive regions in the image. After that, combining the morphology-gradient operator with OTSU segmentation algorithm, it constructs multi-structural element to detect the edges in sensitive regions. The presented algorithm is applied to the asphalt pavement crack detection. The experimental result indicates that compared with other edge detection algorithms, it has not only a good ability to extract the image edge, but also a strong ability to suppress the noise in the image. Moreover, it can provide good image processing effect and processing speed.[著者文摘]

基于小波变换和微分算子的图像边缘检测

The Image Edge Detection Based on Wavelet Transform and Differential Operator

崔夏荣[1] 陆爱萍[2]

本文提出了一种检测彩色图像边缘的新方法，该方法先把彩色图像分解为红、绿、蓝三基色图像，然后利用小波变换对三基色图像分别进行多尺度的分解，提取其低频系数，用微分算子检测其图像边缘，最后利用多通道信息融合的办法将三基色的图像边缘融合在一起。通过一个实验说明了该方法的有效性。[著者文摘]

In this paper, a new method is proposed to detect the edge of color image. Firstly, color image is separated to the three base color images of redness, greenness and blueness. Then the three base color images are decomposed respectively by wavelet transform with more scale, low frequency coefficients are extracted, and image edge is detected by differential operator. Finally, the edges of three base color images are integrated by the approach of the muhichannel information fusion. Validity to the method is also provided through a practical example.[著者文摘]

基于Simulink的图像及视频边缘检测

Image/Video Edge Detection Based on Simulink

赵雪英

边缘检测是图像分割、模式识别等图像分析领域十分重要的基础课题。MATLAB的Simulink是进行动态系统的模型化、仿真和分析工具。阐述了传统的边缘检测方法一微分算子法，应用MATLAB 2007 Simulink的视频和图像工具集进行图像及视频边缘检测，给出仿真模型并进行仿真结果比较。[著者文摘]

The edge detection is the most important basic topic in the image segmentation, pattern recognition, etc. Simulink is the software for modeling, simulating, and analyzing dynamic systems. Traditional method of edge detection-differential operator method is introduced. The image and video processing blockset of Simulink is used to detect the edge of image and video, the simulation models are given and the results are compared.[著者文摘]