Three-dimensional models are brought together on the basis of the confidence parameters

摘要

System for the combination of a plurality of three-dimensional models, each of which consists of images of a camera with a different viewing angle have been produced, comprising:one or more processors;a memory, which with the one or more processors is coupled;a voxel - determination module, which is configured such that it comprises a plurality of voxels in a three-dimensional space determined;a distance test module which is configured so that it is for the respective voxels in the plurality of voxels a plurality of distance values are determined, wherein each distance value a distance from the voxels to form a three-dimensional model of the plurality of the three-dimensional models are along a perspective view of a camera module, that is used, in order to construct the three-dimensional model, wherein the distance test module in one or more processors is implemented;a confidence measure module which is configured such that, when (i), a first distance value of the plurality of the distance values indicates that the distance between the voxel and the three-dimensional model, which corresponds to the first spacing value, within a threshold value, and (ii) a second spacing value of the plurality of the distance values indicates that the voxel outside of the threshold value, and by means of the three-dimensional value, which corresponds to the second distance value:Determination of a first confidence parameter, which indicates a degree, in which the perspective of the camera model of the three-dimensional model, which corresponds to the first spacing value, is aligned in the direction of the voxel; andDetermination of a second confidence parameter, which indicates a degree, in which the perspective of the camera model of the three-dimensional model, which corresponds to the second distance value, is aligned in the direction of the voxel; anda point selection module, which is configured such that it determines whether a point which corresponds to the voxels, in a three-dimensional model, which is guided at least partially based on the first and second confidence parameter is included, so that, if the second confidence measure by a previously determined degree greater than the first degree of confidence is, the voxel is not in the form of three-dimensional model is included.