安检系统中X射线透射图像处理方法研究

发布时间：2018-03-26 09:03

本文选题：安检　切入点：X射线图像　出处：《沈阳大学》2015年硕士论文

【摘要】：随着人们生活水平的提高和近几年电子商务的迅速发展,物流行业中邮路得到了很大的发展,邮路的安全检查关系着民众的人身和财产的安全,更受到了广泛的关注。然而在实际的安全检查中,由于相关技术还不够完善,经常会出现错检和漏检等情况,因此如何提高危险品检测的效率和准确度,减少意外或危险事故的发生有着非常重要的研究价值和现实意义。本文的研究对象是用于邮路安全检查的X射线安检机产生的X射线透射图像,通过对这些X射线图像的处理方法的研究,结合一些危险品的识别原理,欲提高安检机的危险品检测准确率。本文对X射线安检机的构成及工作原理和X射线图像的成像原理及特性进行了详细的理论研究。根据X射线安检机所呈现的X射线图像的颜色特征,利用Ostu方法和基于HSI色彩空间的分割方法对X射线图像进行了危险品图像的分割,将疑似危险品区域的图像分割出来。对广义Hough变换理论进行了研究分析,并用广义Hough变换进行危险品图像的匹配。为了提高对旋转图像的匹配识别率,对广义Hough变换进行改进,提出了一种将数学统计理论和广义Hough变换相结合的改进方法,引入一个距离阈值?,增加了一个匹配标准因子,有效提高了危险品匹配的精度。在MATLAB环境下,选取若干类危险品的图像进行危险品检测的实验验证。实验结果表明,本文提出的改进方法在匹配旋转的图像方面要比传统广义Hough变换有着更高的匹配精度,使危险品检测的准确度有所提高。验证了本文提出的方法的有效性。
[Abstract]:With the improvement of people's living standards and the rapid development of electronic commerce in recent years, postal routes in the logistics industry have been greatly developed. The security inspection of postal routes is related to the safety of people's life and property. However, in the actual safety inspection, because the related technology is not perfect, there will often be false check and missed inspection, so how to improve the efficiency and accuracy of dangerous goods detection, Reducing the occurrence of accidents or dangerous accidents has very important research value and practical significance. By studying the processing methods of these X-ray images and combining the principles of some dangerous goods recognition, In order to improve the accuracy of dangerous goods detection, this paper makes a detailed theoretical study on the composition and working principle of the X-ray security machine and the imaging principle and characteristics of the X-ray image. The color characteristics of X-ray images, Using the Ostu method and the segmentation method based on HSI color space, the dangerous goods image of X-ray image is segmented, and the image of suspected dangerous goods region is segmented. The generalized Hough transform theory is studied and analyzed. In order to improve the matching rate of rotating image, the generalized Hough transform is improved, and an improved method combining mathematical statistics theory with generalized Hough transform is proposed. Introduce a distance threshold? In this paper, a matching standard factor is added to improve the accuracy of dangerous goods matching effectively. In MATLAB environment, some dangerous goods images are selected for dangerous goods detection. The experimental results show that, The improved method proposed in this paper has higher matching accuracy than the traditional generalized Hough transform in matching rotating images, which improves the accuracy of dangerous goods detection, and verifies the effectiveness of the proposed method.
【学位授予单位】：沈阳大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.41

【参考文献】