使用由方差和平均梯度构造的新的评价因子——小波邻域信息量作为融合规则选取小波高频系数。
For the image high-frequency part, use the new evaluation factor - wavelet neighborhood information to choose the ultimate wavelet high-frequency coefficients.
去噪算法的基本思想是平均，关键点在于如何使图像得到平滑的同时，细节或是高频部分予以保留。
The basic idea of denoising is what is called average, so the main point is how to deal with smoothing while preserving details or high frequency parts.
在8个汉语高频多义次的测试中取得了平均准确率为84.52%的较好的效果，验证了该方法的有效性。
The experiments on 8 ambiguous words we get the result of average accuracy 84.52%, which show that the method in this paper is very good.
应用推荐