基于轮廓检测与目标定位的DICOM图像分析测量系统

DICOM Image Measurement System Based on Contour Detection and Target Location

针对由于DICOM图像的复杂性导致图像目标测量不准确的问题,文章在三维图像数据的基础上,开发了相应的目标定位和距离测量算法,设计了一种利用轮廓检测和目标定位的方法来分析和测量DICOM图像。首先,通过CT设备获取原始图像数据,并分析原始数据的结构特征,使用开源图像库ITK解析DICOM图像,将其读取到系统软件的内存中,并在自主开发的软件界面上显示。其次,对DICOM图像进行轮廓检测,确定目标所在帧。然后,结合分水岭算法和水平集分割算法,设计了定位模型,完成目标的定位和测量,并输出定位的图像结果。最后,利用QT开发平台实现了测量和定位算法以及相应的软件功能。测试结果表明:与当前的DICOM图像测量技术方案相比,文章所提方法具有更高的测量和定位准确度。

In response to the issue of inaccurate measurement of image targets due to the complexity of DICOM images,this paper had developed corresponding algorithms for target localization and distance measurement based on three-dimensional image data.This had led to the design of a method that utilizes contour detection and target localization to analyze and measure DICOM images.Firstly,the raw image data was obtained through a CT device,and the structural features of the raw data are analyzed.The open-source image library ITK was used to parse DICOM images,reading them into the memory of the system software and displaying them on a self-developed software interface.Next,contour detection was performed on the DICOM images to identify the frames containing the targets.Subsequently,a localization model was designed by combining the watershed algorithm and the level set segmentation algorithm to achieve target localization and measurement,with the output of localized image results.Finally,the measurement and localization algorithms,along with the corresponding software functionality,were implemented using the QT development platform.Test results demonstrated that the method proposed in this paper exhibited higher measurement and localization accuracy compared to current DICOM image measurement techniques.