MFOC-CliqueNet: A CliqueNet-Based Optimal Combination of Multidimensional Features Classification Method for Large-Scale Laser Point Clouds

Table 4

Weighted combination results of multidimensional features (%).

3D : 2D	Class
	Pole	Vegetation	Wire	Ground	Facade	OA
0.9 : 0.1	0.028	0.965	0.190	0.979	0.716	0.9822
0.8 : 0.2	0.032	0.902	0.299	0.962	0.475	0.9762
0.7 : 0.3	0.083	0.837	0.555	0.935	0.474	0.9744
0.6 : 0.4	0.039	0.812	0.407	0.971	0.594	0.9801
0.4 : 0.6	0.025	0.803	0.232	0.913	0.853	0.9724
0.3 : 0.7	0.135	0.677	0.299	0.982	0.842	0.9783
0.2 : 0.8	0.023	0.777	0.678	0.981	0.615	0.9750
0.1 : 0.9	0.076	0.519	0.276	0.977	0.907	0.9690

The bold values mean optimal overall classification results after weighted combination of multidimensional features of point cloud.