Registration Method Of Three-dimensional Laser Point Cloud And High-definition Image Based Unified Scale

Three-dimensional laser point cloud can express the three-dimensional spatial coordinates and intensity of the target object,but it does not have real color,texture and other information;However,high-definition images are rich in spectrum,color and texture information,but do not include the elevation,position and intensity of the target object.The registration of 3D laser point cloud and high-definition image can make up for the deficiency of single data,realize the complementary advantages of the two,and make it play a greater role in more research and application fields.Because there are great differences between the two kinds of data in terms of acquisition method,organization structure and coordinate system,the following contents are studied when performing registration:(1)Image point cloud scale estimation with three-dimensional laser point cloud as reference.Three-dimensional laser point cloud and image point cloud are two kinds of heterogeneous data,and there are scale differences between them.In order to make them unified in scale,based on the analysis and reproduction of Scale ratio ICP algorithm,a model grid resolution method is proposed to solve the problem.Because3 D laser point cloud can reflect the actual scale of objects,the proposed method is used to estimate the scale of image point cloud with 3D laser point cloud as reference.(2)Registration of three-dimensional laser point cloud and image point cloud.In this study,the registration between 3D laser point cloud and HD(high-definition,HD)image is transformed into 3D laser point cloud and image point cloud.Therefore,on the basis of unified scale,the registration method of 4PCS(4-Points Congruent Sets Algorithm,4PCS)and ICP(Iterative Closest Point,ICP)is used to realize the registration between 3D laser point cloud and image point cloud after scale conversion.(3)Evaluating of scale calculation results and registration accuracy.The relative error and root mean square error are used to quantitatively evaluate the scale calculation results and registration accuracy respectively,and the relationship between them is analyzed.On this basis,the scale calculation results and registration results are dynamically adjusted to achieve the best scale calculation results and the best registration effect.In order to verify the feasibility of the overall technical process and the accuracy and robustness of the proposed method,the relevant data provided by other research teams and the actual data collected by our research group were used for experimental analysis.The experimental results show that the proposed model grid resolution method can complete the initial scale estimation of point cloud,and can determine the optimal scale by dynamic adjustment method Based on the scale conversion of image point cloud by using the scale calculation results,the registration of three-dimensional laser point cloud and image point cloud can be realized by combining 4PCS algorithm with ICP algorithm Relative error and root mean square error evaluation index can quantitatively evaluate the scale calculation results and registration accuracy,and play an important role in determining the optimal scale and registration results.