煤矿井下输送带裂纹原始图像含有各种噪声,在对图像处理前,需先对原始图像去噪.因此,提出一种改进的小波变换阈值去噪算法,该算法综合了分数阶微分和小波分解的算法,对目标图像依次进行多层次、多尺度的剖分,并通过含有8个对称方向上的分数阶微分掩膜算子,依次将目标图像中含有的高频和低频成份重构后再叠加.实验仿真结果表明,该算法克服了小波硬阈值去噪法存在的图像视觉失真和小波软阈值去噪法存在的图像边缘模糊等缺点,获得较好的图像去噪效果.%The coal mine conveyor belt crack the original image contains all kinds of noise, the image processing, first of all, the original image denoising.In this paper, an improved wavelet threshold denoising algorithm, this algorithm integrated the fractional order differential and the algorithm of wavelet decomposition, the target image of multi-level and multi-scale decomposition, and contains eight symmetry directions of fractional order differential mask operator, in turn, the target image contains high frequency and low frequency components after reconstruction, stack again.Experimental simulation results show that this algorithm overcomes the hard threshold wavelet denoising method of image visual distortion and wavelet soft threshold denoising method of fuzzy image edge and other shortcomings, get a better image denoising effect.
展开▼
