Research Article
How Many Bedrooms Do You Need? A Real-Estate Recommender System from Architectural Floor Plan Images
Algorithm 2
Room segmentation and decor items retrieval.
| Input: | | coordinate point of an independent room image | | x-axis image size of an independent room | | y-axis image size of an independent room | | Output: | | number of independent rooms | (1) | ; | (2) | ; | (3) | While = 1 indicates that a new area has been found | (4) | do | (5) | ; | (6) | ; | (7) | fordo | (8) | fordo | (9) | if pixel intensity of 255 indicates white | (10) | then | (11) | ; | (12) | ; | (13) | break | (14) | ; | (15) | While Q = 1 indicates there may be missing points | (16) | do | (17) | ; | (18) | fordo | (19) | fordo | (20) | Ifthen | (21) | Convert the pixel values to 255 in the top, bottom, left, and right directions of points into | (22) | This process is terminated when it encounters 0. | (23) | if a pixel is converted then | (24) | ; | (25) | else ifthen | (26) | Determine whether there is T before 0 is encountered in the four directions of point . | (27) | | (28) | for do | (29) | if These two cases are not considered as a room | (30) | then | (31) | continue; | (32) | else | (33) | Convert all to 0 and convert other values to 255. | (34) | Stack the decor items on the respective detected rooms. |
|