基于小波变换的目标检测关键算法研究  

Core Algorithms of Target Detection based on Wavelet Transform

作　　者：曹量[1] 李太君[1] 

机构地区：[1]海南大学信息科学技术学院,海南海口570228

出　　处：《通信技术》2011年第9期77-79,82,共4页Communications Technology

基　　金：2010年海南省工程技术研究中心专项基金项目(No.2060499)

摘　　要：图像去噪和边缘检测都是目标检测中至关重要的过程。结合半软阈值法和比例萎缩法,提出了一种改进的小波局部自适应阈值图像去噪算法,给出了一种高斯小波模极大值算法进行图像边缘检测。实验证明,去噪算法优于传统的半软阈值算法,较好地保留了图像的细节信息。边缘检测算法与经典的Canny算子相比得到了更多细节信息,使图像变的更真实。Image denoising and edge detection are the critical processes for target detection.In combination of semi-soft threshold method and rate-contraction method,an improved local adaptive wavelet thresholding image denoising algorithm is proposed,and a Gaussian wavelet modulus maxima algorithm for image edge detection is given.Experimental results show that the denoising algorithm is superior to the traditional semi-soft threshold method,and could fairly retain the image details,and that the edge detection algorithm could achieve more detailed information as compared with classical Canny operator,thus the image has more sense of reality.

关 键 词：小波变换 图像去噪 边缘检测 自适应阈值 模极大值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]