Multi-view can provide more object information than single view and are less susceptible to noise interference. But in the feature matching process, excessive parallax in multi-view can lead to mismatches. And it is difficult to extract features for weakly textured area, which will causes the reconstructed model contain holes. Here, the method for the reconstruction of 3D structural semantic points with multiple camera views is presented. We take the advantage of the multi-view method and 3D feature points to reduce mismatching in the feature matching process. The constraints that provided by structure semantics points are related to object and restrict the distribution of points around object, which can improve the reconstructed model. Besides, the model with 3D feature points can be optimized using semantics and distance information to fill holes and remove noise. The experiment uses eight cameras to test method. The results show that our method can be effective for mismatching and holes. The experiment results prove that our method is effective.
In recent years, 3D display technique is one of the emerging technology and gradually becomes accessible to a broader audience. However, because of the traditional 3D reconstruction method is limited by the number of the feature found in the image, the resolution of the generated 3D model is not high enough for 3D display. A new system is purposed, in which we consider the vertical and horizontal disparity between images, and the optical flow is used to replace the feature matching segment, so that more points can be pushed into the reconstruction process for improving the resolution of the models. Experimental results prove that the resolution of the models can be enhanced effectively. The details of the model are preserved, and the holes in the weak texture region are successfully filled.