论文部分内容阅读
为了以更高的精度进行路面裂缝类别判断,基于3D数据扫描技术,提出双相扫描路面裂缝初识别方法;在此基础上,采用区域生长算法进行裂缝进一步识别;使用线性拟合与双向扫描求取交叉点的方法进行网状裂缝类型判断;提出了裂缝区域的矩形模型,通过使用距离以及角度判断的方法由断裂裂缝模型获得完整裂缝矩形模型,并基于该矩形模型进行了线性裂缝区域信息的表征及类型的判定;利用准确率-召回率曲线进行算法精确度定量评价。试验结果表明:初识别的裂缝既包含裂缝信息又包含非裂缝信息;区域生长算法可进行裂缝进一步识别;线性裂缝的矩形模型可用来进行线性裂缝区域信息表征;PR曲线可用于定量分析算法精确度。
In order to judge the type of pavement cracks with higher accuracy, based on 3D data scanning technology, the method of initial identification of pavement cracks in dual-phase scanning is proposed. On the basis of this method, the area growth algorithm is used to further identify cracks. Using linear fitting and bidirectional scanning The intersection point method is used to judge the type of the mesh fracture. A rectangular model of the fracture area is proposed. The complete fracture rectangle model is obtained from the fracture fracture model by using distance and angle judgment. Based on the rectangular model, Characterization and type of judgment; the use of accuracy rate - recall curve for quantitative evaluation of the accuracy of the algorithm. The experimental results show that the initially identified fractures contain both fracture information and non-fracture information. The regional growth algorithm can further identify the fracture. The rectangular fracture model can be used to characterize the information of the linear fracture region. The PR curve can be used to quantitatively analyze the accuracy of the algorithm .