Fig. 1. Framework of proposed method for the extraction of indoor objects point cloud

Fig. 2. Point O and its adjacent points

Fig. 3. Distribution of adjacent points. (a) Neighbors bias to one side; (b) neighbors distribute evenly

Fig. 4. Extraction of wall boundary. (a) Original point cloud of wall; (b) wall boundary

Fig. 5. Magnitude distribution of density function of wall

Fig. 6. Extraction of wall according to density cluster of wall point cloud. (a) Original point cloud; (b) extracted wall point cloud

Fig. 7. Density curves. (a) Density curve of floor; (b) density curve of ceiling

Fig. 8. Extracted ceiling and floor point cloud. (a) Original point clouds of ceiling and floor after removing walls; (b) extracted ceiling and floor point clouds

Fig. 9. Cluster centers determined by different methods. (a) Center points determined by CFDP method; (b) center points determined by proposed method

Fig. 10. Different indoor scenes (apartment, bedroom, boardroom, lobby, and loft)

Fig. 11. Different colors show extracted walls, ceilings, and floors of apartment and bedroom. (a) Extracted walls, ceilings, and floors of apartment; (b) extracted wall, ceiling, and floor of bedroom

Fig. 12. Different objects in the apartment are extracted by different methods. (a) CFDP method; (b) DPC method; (c) proposed method

Fig. 13. Different objects in the bedroom are extracted by different methods. (a) CFDP method; (b) DPC method; (c) proposed method

Fig. 14. Different objects in the loft are extracted by different methods. (a) CFDP method; (b) DPC method; (c) proposed method

Fig. 15. Different objects extracted by CFDP, DPC, and proposed method, and displayed with different colors. (a)(b) Boardroom and lobby; (c)(d) CFDP clustering results; (e)(f) DPC clustering results; (g)(h) clustering results of proposed method

Fig. 16. Three types of indoor point clouds. (a) Almost all the objects in the room are independent and not close against together; (b) a few objects in the room close against together; (c) many objects in the room close against together

Fig. 17. Extraction results of wall, ceiling, and floor. (a)(c) Actual wall, ceiling, and floor; (d)(f)extracted wall, ceiling, and floor by proposed method

Fig. 18. Extracted three types of indoor point clouds by proposed method. (a)(c)(e) Actual objects; (b)(d)(f) different objects extracted by proposed method

Fig. 19. Overlap degree of different objects. (a) The first type of room; (b) the second type of room; (c) the third type of room

Fig. 20. Extraction results of different objects in ten offices

Fig. 21. Different objects in two lounges are extracted by proposed method. (a) Extraction result of the first lounge; (b) extraction result of the second lounge

Fig. 22. Extraction results of three hallways by proposed method. (a) The first hallway; (b) the second hallway; (c) the third hallway

Fig. 23. Extraction results of two water closets and two storage rooms. (a) Water closet 1; (b) water closet 2; (c) storage room 1; (d) storage room 2