Urban scene reconstruction method and system

Abstract

The invention discloses an urban scene reconstruction method and system, and relates to the technical field of image processing. The method comprises the steps: firstly extracting sufficient and reliable candidate space planes from the spatial points of an urban scene image through employing a spatial plane fitting method based on connected domain detection; secondly distributing the candidate space planes for all segmented regions obtained through the segmenting of each urban scene image under the frame of minimum MRF energy. The method can iron out the defects that a conventional method is poor in reliability and not complete in reconstruction scene, and is higher in calculating efficiency.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a method and system for reconstructing an urban scene. Background technique [0002] 3D urban scene models are widely used in urban planning, virtual tourism, and cultural heritage protection, and accurate 3D modeling of urban scenes has important research significance. In practice, although laser scanning equipment can be used to obtain high-precision three-dimensional information, its application range is limited due to the high cost of scanning and the limitation of scanning range. Relatively speaking, the image-based 3D scene reconstruction method is still a common method in urban scene modeling due to its low cost and convenient data acquisition. [0003] However, it is still a challenging problem to effectively obtain a complete urban scene model using image information due to factors such as urban scene illumination changes, perspective distortion, weak texture a...

AI Technical Summary

Problems solved by technology

[0003] However, it is still a challenging problem to effectively obtain a complete urban scene model using image information due to factors such as urban scene illumination changes, perspective distortion, weak texture and repetitive texture regions.

Among many related algorithms, although reconstruction algorithms based on specific scene models or regions (such as superpixels) can overcome the shortcomings of pixel-level reconstruction algorithms that are difficult to obtain a complete scene structure to a certain extent, there are also the following problems that lead to its Low reliability and efficiency: (1) The scene model used is too simple (such as the Manhattan scene model with only 3 orthogonal directions) or the structure of the scene is too limited (such as pre-specifying several scene directions), which leads to reconstruction There are large deviations in the results; (2) The candidate plane set used for scene complete structure inference lacks completeness (that is, only part of the plane in the scene is included), which affects the reliability of the inference results; (3) The non-reconstructed Areas (such as the sky, the ground) lack effective detection and filtering mechanisms, and the overall reconstruction efficiency is therefore greatly affected

Images