基于图像边缘特征的计算机图像检测算法研究

Research on Computer Image Detection Algorithm Based on Image Edge Feature

计算机影像在对真实影像进行仿真时,对影像内容的一致性进行了严格的控制,但对于一些不变的部位,特别是边界部位,还是会有较大的差异;它也许难以与大自然的景象相比,而且难以掌控。针对目前计算机图像在边界构建上存在的不足,提出了一种利用小波高阶特性对边界进行刻画的新方法,并将其与预报错误图像中的高阶特性相融合;在此基础上,本文介绍了一种以边缘提取为基础的计算机图像提取方法。

Computer images in the process of simulating natural images,well control the consistency of the image content,but there may be obvious differences in some fixed positions,especially in the edge of these positions,may be difficult to compare with the natural image,is also difficult to control.In this paper,based on the defect of computer image in the construction of edge,starting from the edge of the image,using wavelet high-order features to describe the edge difference of the image,and combining the high-order features of the prediction error image wavelet transform,a computer image detection algorithm based on edge detection is proposed.