基于Zadeh-x变换的法医学底层图像挖掘

Lower layer image mining in forensic medicine on the zadeh-x transformation

目的:利用Zadeh-x变换对法医学中在恶劣的条件下获得的、被强背景干扰掩埋人眼无法辨认而实际却存在的图像信息进行底层图像挖掘,为法医学提供尽可能清晰的图像特征信息作为物证。方法:利用VC++编程实现Zadeh-x变换(Zadeh-x transformation)对法医学图像资料中背景与目标图像信息有差异的底层图像进行挖掘。结果:通过与传统算法如直方图均衡化比较,实现了对暗背景、电泳图、车标和衣服等图像中隐藏信息的特征提取。结论:Zadeh-x变换的方法具有较强的提取图像隐藏信息的能力,其处理结果明显优于直方图均衡化,甚至可以处理用直方图均衡化无法处理的图像信息。

Objective：to offer details as clear as possible to be material evidence for Forensic Medicine from the images which were obtained in bad conditions burying in the strong back and can’t be identified by human vision . Methods：using zadeh-x transformation to mine useful information hiding in the dark back. We implement the zadeh-x transformation by VC＋＋ programming on computer,achieving the lower layer image mining in Forensic medicine pictures whose foreground is different from its background. Results：we have realized the lower layer image mining in pictures burying in black ground,electrophoresis graph,vehicle brand and clothing character. Moreover, we have compared the result with the picture dealing by histogram equipoise. Conclusion：The lower layer image mining in Forensic Medicine has been achieved by VC＋＋ programming in this article. Its result is distinctly better than the result dealing by histogram equipoise. Further more,it can mine some information that histogram equipoise can’t do.