基于改进的离散傅里叶变换图像分割算法研究  

Discrete Fourier Transform Based on Improved Algorithm for Image Segmentation

作　　者：张建梅[1] 孙志田[1] 李香玲[1] 

机构地区：[1]河北建筑工程学院,河北张家口075024

出　　处：《计算机仿真》2012年第3期300-302,315,共4页Computer Simulation

摘　　要：研究图像分割特征提取优化问题。由于外界信号噪声等问题而引起图像分割分辨率低,清晰度不高,提取图像的主要特征目标是图像分割中关键的技术,针对传统的图像特征提取分割算法无法完成对图像关键特征信息适度提取,另外图像分割计算复杂,为了有效的对图像进行分割,提出了一种改进的离散傅里叶变换的图像分割算法。采用傅里叶变换算法对图像中感兴趣的区域进行分割出来后,对各个分割区域特点进行描述并组成一定的结构,从而获得最优图像分割结果。仿真结果表明,改进的算法可以有效地提取复杂图像区域中的特征信息,分割效果明显,提高了图像分割的分辨率和清晰度。Due to the external signal noise and other problems caused by the low resolution image segmentation,extraction of image clarity is not high.In order to effectively segment image,we put forward an improved discrete Fourier transform algorithm for image segmentation.FFT algorithm was used for the segmentation of interested regions in the image,and the segmented regions were described and composed with a certain structure,thereby obtaining the optimal image segmentation results.The simulation results show that,the improved algorithm can effectively extract the feature information in the complex image region segmentation and improve the image resolution and clarity.

关 键 词：图像分割 傅里叶变换 特征提取 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]