【摘要】 采用小波图像降噪方法容易去除图像的边缘细节特征,导致关键信息丢失,提出一种有效保留图像细节的自适应图像消噪方法。采用局部二值拟合方法对图像进行二值化处理,提取图像的边缘轮廓特征并进行多尺度分割,采用局部区域梯度模特征匹配方法调整邻域的大小,对图像进行全局分割,从而有效保留了图像的细节特征,在此基础上利用变尺度的小波降噪方法实现图像自适应降噪。仿真结果表明,采用该方法进行图像消噪能有效保留图像细节特征,图像消噪输出的峰值信噪比较高,说明图像质量较好,性能优越。更多还原

【Abstract】 Using wavelet image denoising method is easy to remove the image edge details characteristics, lead to the loss of critical information, put forward new adaptive method for image denoising effectively preserving image details. The local two value fitting method for image binarization processing, extract image edge contour feature and multi-scale segmentation, using local regional gradient feature matching method to adjust the neighborhood size on the overall image segmentation, so as to effectively preserve the image details, on the basis of using the wavelet denoising method to achieve image variable scale adaptive denoising. The simulation results show that using the method of image denoising can effectively preserve image detail features, image denoising output the high peak signal to noise ratio, indicating a better image quality,superior performance.更多还原