论文部分内容阅读
本文针对X光片数字化时所引出的噪声以及采用普通图像增强方法信噪比低的问题,提出一种自适应邻域的特征增强方法。考虑到X光片中特征图形的形状、尺寸各不相同的特点,采用了领域形状可变的处理方法,并且利用象素在对比度域内的几种特性,有选择、有针对性的设计出对比度增强函数,使处理后图象的特征得到最大的增强,而噪声得到相应的抑制。在选择对比度增强函数时,可以根据具体需要,结合一些先验的知识,使增强后的图象对比度、清晰度提高,更具有辅助诊断作用。
In this paper, aiming at the problem of the noise induced by the digitization of X-ray film and the low signal-to-noise ratio (SNR) using the normal image enhancement method, a feature enhancement method of adaptive neighborhood is proposed. Taking into account the characteristics of the X-ray images of the shape and size of the different characteristics of the field using a variable shape processing methods, and the use of pixels in the contrast of several characteristics of the domain, selective, targeted design contrast The enhancement function maximizes the features of the processed image, while the noise is suppressed accordingly. When choosing the contrast enhancement function, according to the specific needs, combined with some prior knowledge, the enhanced image contrast, sharpness, more auxiliary diagnosis.