【作者】 马社祥； 刘铁根； 陈鲤江；

【Author】 MA Shexiang LIU Tiegen Chen Lijiang(College of Precision Instrument and Optoelectronics Engineering,Tianjin University, Tianjin 300072Department of Optoelectronics Information and Electronics Engineering, Tianjin University of Technology, Tianjin 300191)

【机构】 天津大学精密仪器与光电子工程学院；

【摘要】 图象的边缘检测在图像分析与理解、图像恢复和增强、图像去噪、图像压缩编码等具有十分重要的作用.为了检测小结构的轮廓以及大目标的边缘,将多尺度边缘检测与小波变换有机的结合起来,利用小波分析方法来研究信号的多尺度边缘特征,即多尺度边缘检测方法。本文针对图像信号的多尺度边缘检测和重构问题,利用二进小波变换的多尺度分析特性,定义了图像在水平和垂直方向的多尺度边缘。同时,利用三次样条插值算法,提出了一种由二进小波变换在水平和垂直方向的极值重构图像信号的算。实验结果显示本文的多尺度重建方法与著名的交替投影算法相比,算法复杂度低,图像重建速度提高了20倍,而且重建图像质量较好,其峰值信噪比提高了1dB以上。更多还原

【Abstract】 The sharp variation points of image contain most important information for image processing, such as image reconstruction, image analysis, image recognition, image enhancement and image compression coding. These points are generally located at the boundaries of the image. In order to detect the boundaries of different structure, multiscale edges detection was introduced. Multiscale edges detection of image signals and the image reconstruction are the major problems for image processing. The paper describes the multiscale directional edges of image with dyadic wavelet transform. Meanwhile, using cubic spline interpolation algorithm, an image reconstruction scheme with the horizontal direction and vertical direction extrema of dyadic wavelet transform is introduced. The experimental results show that the algorithm is effective. Compared with previous algorithm, it has a low complexity, and has a high quality of image reconstruction.更多还原