Method for reconstructing three-dimensional shape of cumulus based on single natural image

Abstract

The invention discloses a method for reconstructing the three-dimensional shape of a cumulus based on a single natural image, and belongs to the field of computer graphics. The method can automatically construct the three-dimensional shape and surface details of the cumulus from the single natural image. The method comprises the steps of firstly, combining a shape from shading algorithm with cloudsurface shading information to obtain a shading constraint; secondly, calculating a boundary constraint of the cumulus by using the uniformity of contour and self-shading information of the cumulus;thirdly, constraining illumination parameters by using a two-order spherical harmonic function, and then calculating depth information of the cumulus; and finally, generating a cumulus mesh model, sampling inside the cumulus, forming a particle model of the cumulus and drawing the particle model. By introducing the shading constraint and the boundary constraint to guide the construction of a cumulus scene, the cumulus shape and detail information consistent with the image can be effectively recovered; and the obtained cumulus model is natural and rich in details.

Description

technical field [0001] The invention belongs to the field of computer graphics, in particular to the field of cloud modeling, in particular to a method for reconstructing the three-dimensional shape of cumulus clouds based on a single natural image. Background technique [0002] As a common natural phenomenon, clouds have ever-changing shapes, and the process of formation, development and dissipation is extremely complex. Cloud modeling has always been a challenging task. Classical cloud modeling methods are roughly divided into two categories, namely, process-based methods and physical simulation-based methods. The former mainly uses fractal theory and noise texture to model the shape of clouds, and can build realistic clouds, but the modeling process relies on frequent parameter adjustments. In order to reduce or even avoid parameter setting, researchers began to try a method based on physics simulation, that is, to simulate the cloud generation process by simulating simp...

AI Technical Summary

Problems solved by technology

However, their method is not suitable for modeling cumulus

On the one hand, it is difficult to record cumulus information in low-resolution satellite images, and on the other hand, because the method relies on parameter settings, the error in geometric thickness is too large, even exceeding the typical size of cumulus clouds

Images