【作者】 李绪武；

【导师】 刘飞；

【作者基本信息】 重庆大学 ， 机械制造及其自动化， 2013， 博士

【摘要】 基于三维扫描工程建模的面部整形技术属机械制造技术与医学工程技术的交叉领域，是当前研究的一个热点问题；能有效避免传统方法中依靠医生的主观经验完成临床治疗手术的不确定性。三维扫描工程建模的点云数据处理是该项技术的关键技术。本文针对现有研究的不足，对基于点云数据的缺陷修补方法、基于点云数据数学模型求解的脸部整形面积获取方法、基于点云数据的正畸牙齿模型空间移动精确距离分析方法等面部整形点云数据处理中的关键方法进行了研究。主要包括以下内容：①针对面部整形点云数据缺陷修补问题，提出了通过多幅二维图片对三维点云数据进行缺陷修补的方法。该方法以ATOS扫描设备所获取的点云数据为研究对象，首先对扫描得到的点云数据进行预处理，提取二维图片与三维点云数据中非编码点的位置坐标，同时提取三维点云数据中缺陷点云数据的边界，接着通过非编码点建立二维图片与三维点云数据之间的对应关系，并求出焦点位置，然后建立焦点与所要修补的非编码点之间的空间直线方程，最后通过计算多条直线交点来确定需要修补的缺陷点云数据位置坐标。②针对大面积烧伤、创伤的自体或体外皮源移植中的面部整形面积计算问题，提出了一种基于点云数学模型求解的脸部整形面积获取方法。该方法首先进行面部扫描，然后对点云数据进行预处理，接着对预处理后的点云数据采用对称化处理，得出对称前后两个点云数据的色彩对比图CDM（Color Difference Maps）,通过CDM图得到脸部整形面积区域的切片数学模型，将该切片模型直接进行求解从而最后获取脸部整形面积。该方法直接对点云数据进行操作，可以准确、高效地完成手术面积计算。最后通过对比分析表明这种方法在脸部整形中是行之有效的。③针对正畸矫治过程中牙齿空间移动距离值的精确测量问题，提出了一种基于Typodont水域实验的三维空间移动距离测量的新方法。该方法首先通过Typodont水域实验得到牙齿模型的三维数据，然后在迭代最近点ICP（IterativeClosest Point）匹配算法的基础上采用VICP算法对采样点云数据进行二次虚拟匹配，找到与被检测模型相对应点的数据，得到模型移动的三维空间距离值，从而解决了正畸矫治过程中对每个牙齿移动距离以及移动方式的分析。通过模型移动对治疗中牙齿移动距离进行检测，可以对正畸牙进行随时监控疗程，避免牙齿移动过量，起到有效利用支抗的作用。对以上①②③种方法，均给出了实验案例；案例分析结果表明，上述方法是可行和有效的。更多还原

【Abstract】 Facial Plastic surgery technology is based on three-dimensional scanning projectmodeling machinery manufacturing technology and cross-cutting areas of biomedicalengineering technology, it is a hot topic of current research. This technology effectivelyavoied the traditional methods rely on the subjective experience of the doctor tocomplete the uncertainty of clinical surgery, played an objective evaluation of theeffective supporting role. The point cloud data processing engineering modeling ofthree-dimensional scanning is the key technology of the technology. For this lack ofexisting research, our paper researched the defect repair method based on the pointcloud data in the point cloud data processing of the facial plastic surgery, and researchedfacial plastic surgery area acquisition method based on the point cloud data of solvemathematical model, and the exact distance analysis method of the orthodontic toothmodel space based on the point cloud data. For this reason, this paper include thefollowing aspects:①For the facial plastic point cloud data defects, proposed through multipletwo-dimensional images of three-dimensional point cloud data for the defectssupplement, this method is to get through the ATOS scanning equipment scanning pointcloud data for the study. First, preprocessing the scanned point cloud data, extract thenon code point location coordinates of the two-dimensional image andthree-dimensional point cloud data, and extract the defective point cloud data boundaryof3D point cloud data, followed by non code point establish the correspondencerelationship between the dimensional picture and the3D point cloud data, and the focalposition is obtained, then create the space linear equation, finally by calculating theintersection point of most straight line to determine the defective location coordinatesfor the cloud data.②For the problem of a large area burns, trauma autologous transplant or in vitroskin source area of facial plastic computational problems, proposed a facial plasticsurgery area acquisition method based on point cloud mathematical model solution.Firstly, facial scan, preprocessing the point cloud, and then symmetrical treatment onthe basis of point cloud data preprocessing, obtain the CDM (Color Difference Maps),through the CDM diagram obtain the area slice mathematical model of the area of facialplastic surgery, by solving the point cloud data in this area directly, the facial plastic area is acquired. This method operates directly on the point cloud data, can accuratelyand efficiently complete the surgical area. Finally, by comparing the analysis shows thatthis method is effective in facial plastic surgery.③For the accurate measurement of orthodontic appliances during the tooth spacemoving distance value, proposed a new method of measuring the3D space distancebased on Typodont waters experiment, this method through Typodont watersexperiments obtained the3D point cloud data of the dental model, and then use theVICP method conduct secondary virtual match of sampling point cloud data based onthe ICP (Iterative Closest Point) algorithm, find the detection model correspondingpoint data, obtain the model move of the3D space distance value, so as to solve theorthodontic treatment process for each tooth movement distance and move way. Bymonitoring the movement distance of tooth under treatment through model movement,the treatment of orthodontic tooth can be monitored at all times, which preventsexcessive movement of tooth and plays a role of effectively using the anchorage.The above①②③methods are given experimental case; case analysis resultsshow that these methods are feasible and effective.更多还原