数字图像处理技术与实证研究

图片的处理方法多种多样,可以用特征值分解的方法对图片进行信息提取,将此方法与信噪比峰值法、全局质量指标计算法,平均结构化相似度指标的测量法结果作比较,进而证明运用图像特征值分解法的优势。再通过非线性过滤的方法对图片进行进一步的除杂;还可以观察图像频谱中的噪声分布情况,根据这些噪声点的在频率域的分布情况,进而选用合适的滤波器来过滤掉图片中的噪声;为了获得更为清晰的图片,还可以建立模板匹配模型,即:经过有限次的待搜索图与色标模板图的配准过程来完成对目标图片信息的检验和建构,这是此方法较其他方法的优势,也是这篇论文所提倡使用的方法。

The Processing Technology and Empirical Research of the Digital Images

There are many different ways for the image processing. We can use the method of the eigenvalue decomposition to extract the information of the images. We can compare it with the peak signal-to-noise ratio method, universe quality index method and mean structure similarity index measurement method, hence, proving the advantages of using the eigenvalue decomposition method. Then by the method of nonlinear clarification, we can purify the picture further. We can also observe the distribution of the noise in the frequency spectrum. According to the distribution of the noise points in the frequency spectrum, we can choose a proper filter to filter out the noise in the images. In order to get clearer pictures, we also can construct the matching model of the template. That is, we can test and construct the information of the target images with the limited procedures of matching between the search graph and the color code template figure. It is one of the advantages compared with other methods, also the method this paper recommends.