A 3D modeling method and system based on image sequence

Abstract

The invention discloses a three-dimensional modeling method based on image sequences, which includes: image acquisition, photographing device calibration, image preprocessing, feature point extraction, stereo matching and three-dimensional reconstruction, etc., through image acquisition, photographing device calibration, image preprocessing Processing, feature point extraction, and stereo matching to establish the depth image of each image in the image sequence step, and use the depth image in the 3D reconstruction step to determine the weighted average distance of the same 3D feature point under different viewing angles to obtain each 3D feature of the modeling object The three-dimensional space coordinates of the points are used to complete the three-dimensional modeling of the modeling object. The method is simple to operate, has three-dimensional modeling accuracy, and can achieve the effect of image-level three-dimensional modeling. Since it does not need to use a professional scanner, the cost of modeling is reduced, and it can meet the needs of popularization.

Description

technical field [0001] The invention relates to the field of three-dimensional modeling, in particular to a high-precision movie-level three-dimensional modeling method and system. Background technique [0002] At present, most of the 3D models in many games and 3D movies need professionals to use professional modeling tools to make 3D modeling of objects, which will consume a lot of manpower and time. However, using a 3D scanner to extract 3D information of an object has a high equipment price, which will result in high cost of 3D modeling, which does not meet the needs of popularization. [0003] In the scheme of 3D reconstruction of objects, the existing image-based 3D reconstruction methods can be divided into the following categories: (1) single-view 3D modeling, (2) dual-view 3D modeling and (3) multi-view 3D modeling. Because obtaining the depth information of an image based on a single or two images has high requirements on the algorithm, the 3D modeling method bas...

AI Technical Summary

Problems solved by technology

In the existing multi-view 3D modeling scheme, the camera equipment is usually run around the object for dynamic shooting, to obtain a 360-degree image of the required 3D modeling object, and then transmit the image to the computer for 3D modeling, but this To a certain extent, the method has relatively high requirements on the surrounding shooting angle and shooting speed, and due to the uneven distribution of light, it will lead to large errors in 3D modeling, which cannot meet the requirements of high-precision 3D modeling in the existing film and television industry. need

Images