Three-dimensional reconstruction algorithm parallelization method based on GPU cluster

Abstract

The invention discloses a three-dimensional reconstruction algorithm parallelization method based on a GPU cluster, and relates to the technical field of computer vision. The method is based on an SfMalgorithm, researches a drone image three-dimensional reconstruction technology process, and adopts a GPU cluster as a processing platform to solve the problem that drone three-dimensional dense reconstruction processing is time-consuming. Specifically, based on an SFM_MVS correlation theory, a real scene three-dimensional reconstruction correlation process based on a picture sequence is grasped,and meanwhile, an MPI parallel programming technology and a GPU parallel programming technology are adopted to carry out optimization acceleration research work on part of links of the three-dimensional reconstruction process. According to the method, sparse reconstruction algorithm operator replacement is carried out by using a cluster, so that the problems of large data volume and time-consuming calculation of aerial images of an unmanned aerial vehicle are effectively solved; and dense point cloud reconstruction flow in the later stage of three-dimensional reconstruction is effectively accelerated by coarse-grained data parallelism of a dense reconstruction algorithm and fine-grained parallelism and optimization of a dense matching algorithm feature extraction link.

Description

technical field [0001] The invention belongs to the field of GPU clusters and parallel computing, and relates to a method for parallelizing a three-dimensional reconstruction algorithm based on GPU clusters. Background technique [0002] In recent years, UAV technology has developed into a relatively novel remote sensing data acquisition platform that is widely used. Compared with other remote sensing platforms, the UAV platform has the advantages of low cost, easy to use, quick access to the route planning function combined with the ground station, and simple operation for out-of-the-box experiments. Due to the rapid development of UAVs, the 3D reconstruction technology of UAVs can easily obtain the side texture information of objects through UAVs equipped with multi-angle cameras to shoot ground objects, so that more comprehensive 3D information can be obtained. It plays an extremely important role in construction, management and emergency response. [0003] At present, ...

AI Technical Summary

Problems solved by technology

[0004] (1) Compared with the traditional manual modeling method, the use of commercial software for 3D reconstruction has a lot of efficiency improvements, but when performing large-scale operations, the processing cycle is still long;

[0005] (2) The hardware environment requirements of commercial software are also constantly improving with the upgrading of software versions, and the desktop computers used in ordinary offices are often unable to meet the requirements, which also brings a lot of inconvenience to a certain extent;

[0006] (3) Although some commercial software can also support parallel processing under the local area network, its communication module cannot handle the communication problems between the various processing platforms well, and problems such as communication deadlock and missing tasks often occur; and commercial software Closed source, unable to carry out further targeted modification and development

Real-scene 3D reconstruction using open source computer vision algorithms has the characteristics of easy access to source code and high portability. However, with the improvement of people's requirements for point cloud quality, the sparse point cloud obtained by SfM cannot be directly applied, and further 3D reconstruction is required to obtain 3D Dense point cloud, the time-consuming data processing at this time becomes a big problem

Images