基于车载CCD序列影像的建筑立面纹理映射

Reconstructing facade texture of building based on vehicular CCD

在车载LIDAR系统中,CCD相机系统侧重于获取精度的建筑物立面信息,非常适用于城市建筑物模型的快速构建。使用二维最大熵阈值分割对最佳影像进行有效地快速分割,采用Hough算子检测纹理边界,并利用点特征对纹理轮廓和模型轮廓进行匹配纠正。实验证明,该方法简单、快速,适合于大范围城市建筑物立面纹理映射。

As the current research, CCD of Vehicle LIDAR system focuses on the acquisition precision/fast building facade information, which is suitable for city building model. 2D maximum entropy threshold is used to fractionate the image, the texture boundaries are detected by the Hough, and point feature contour and tex- ture are used to match the correct model profile. Experiments show that this method is simple, fast, suitable for a wide range of city building facade texture mapping.