基于帧间相似的序列X光图像的预处理技术

Preprocessing of Serial X-ray Images Based on Inter-frame Similarity Detecting

目的:目前,常规X射线成像技术逐步进入了数字化,为了解决X射线图像数字化处理后的质量问题,本文提出了一种基于动态图像帧间相似性检测的滤波和增强算法。该方法能有效的增强序列X光图像质量,并保证了一定的运算速度,为传统X射线设备的数字化改造提供了基础平台。方法:本文将根据判断序列图像间的关系,采用条件性多帧叠加技术,实现噪声滤除,增强图像信噪比。同时,采用非线性灰度变换曲线,调整图像的对比度。当帧间相似性高时,采用插值技术实现动态图像增强,提高处理速度。结果:采用Matlab软件进行多组图像的仿真实验,图像有效的得到了增强。结论:仿真结果表明,提出的方法能根据相似性关系实现不同程度的噪声抑制、图像增强,有效地提高序列X光图像的显示质量。

Objective： At present,traditional X ray imaging technique enters into digital era gradually. In order to resolve the problems of the X ray images, this paper proposed an filtering and enhancement algorithm, which is based on inter-frame similaxity detecting. This method can enhance the contract of digital x-ray images, and ensure the processing speed ,it can provide the fundamental platform for X ray instruments＇ digitized transformation. Methods： By comparing the similarity between two images, using the technology of conditioned multi-frame superposition, it can filtering noise and increases the signal to noise ratio. Meanwhile, the image contrast can be adjusted by non-linear contrast transformation. This method also chooses different ways to improve SNR of image. The interpolation way is used to enhance dynamic images when they have high similarity, which can improve the speed of processing. Results： Some X- ray images are processed by this algorithm using Matlab software ,imaging enhancement can be realized. Conclusions：The simulation results show that this method can realize noise reducing and contrast enhancement, which can improve the image quality of display devices effectively.