Method of generating 3D house type model by 2D house type model based on camera shooting

Abstract

The invention discloses a method of generating a 3D house type model by a 2D house type model based on camera shooting. The method comprises steps: 1, a 2D house type picture is acquired through camera shooting; 2, an essential matrix reduction method is used for correcting the 2D house type picture acquired in the first step; 3, image enhancement processing is carried out on the2D house type picture after processing of the second step; 4, a contour line of the 2D image after processing of the third step is extracted; 5, a door and window model base is built, matching is carried out according to the contour line extracted in the fourth step, and door and window positions are recognized; 6, a wall structure is stretched to a fixed height position, and generation of a 3D house type model is completed. The method has the following advantages that automatic generation of the 3D house type model from the 2D house type model is realized, the generation method is simple, applicability is universal, the generation method can be embedded in a portable mobile device to form APP application software to be combined with a camera of the portable mobile device, and practicability is provided.

Description

technical field [0001] The invention relates to a method for generating a 3D house type model based on a 2D house type photo, in particular to a method for generating a 3D house type model based on a 2D house type photo captured by a camera. Background technique [0002] The 3D reconstruction based on the 2D building floor plan is mainly divided into two categories according to the different input formats: the 3D reconstruction based on the scanned image and the 3D reconstruction based on the vector floor plan. The method of 3D reconstruction based on scanned images mainly includes: 1) preprocessing scanned images, including denoising, text extraction and vectorization; 2) analyzing vectorized drawings and identifying feature information (walls, doors, windows, balconies, etc.) ;3) 3D reconstruction [0003] How to analyze the vectorized floor plan, correctly parse and extract the geometric and semantic information of building components and the layout information it repres...

AI Technical Summary

Problems solved by technology

[0004] The 3D synthesis method described in the above prior art has the following defects: 1, the quality requirements of the original input picture of the floor plan are too strict; 2, the universality of feature information identification is not good, in the existing methods, each method Basically only for the floor plan of a certain floor plan library, the characteristic information symbol representation of different floor plan libraries is quite different, and there is no mature method that can be generally applied to floor plan maps of various different storehouses

Using the method based on statistical analysis to form a special semantic library through pre-sample can improve this problem to a certain extent, but pre-sample learning still increases the complexity of the algorithm, increases time-consuming, and the effect is not very ideal; 3) For complex apartment types In addition to the necessary architectural originals, walls, doors, windows, and balconies, if there are other furniture, bathroom, decoration, etc. icons in the floor plan, it will increase the difficulty of identification

Images