Surface optimization method and device for reconstructing dynamic objects

Abstract

The invention discloses a surface optimization method and device for reconstructing a dynamic object, wherein, the method comprises the following steps: acquiring a color image sequence and a depth image sequence of the reconstructed object by a single hand-held RGBD camera; according to the color image sequence and the depth image sequence, solving the current camera position for each frame, obtaining the motion of the reconstructed object and the ambient illumination, and updating the current geometric and surface information of the object by the fusion method. A plurality of key frames areselected from the color image sequence according to the definition, and the pixels of the key frames are optimized so that the optimized color image meets the preset conditions on the pixel block andthe original input color image; The geometrical information of the reconstructed object is encrypted, and the optimized color map of each key frame is back projected to the object by using the color map whose resolution satisfies the preset value, so as to obtain the geometry and surface of the reconstructed object. This method optimizes the surface of the object on the dynamic object, obtains clear and accurate surface information, and is widely used.

Description

technical field [0001] The invention relates to the technical fields of computer vision and graphics, in particular to a surface optimization method and device for reconstructing dynamic objects. Background technique [0002] 3D reconstruction is a very important problem in computer vision and iconography, and it has a wide range of applications in the fields of film, animation, virtual reality and augmented reality. The depth camera and the color camera can respectively obtain the depth value and color value corresponding to each pixel, which can be used to reconstruct the geometry and surface of the object, but there are some errors in these data, resulting in the final surface effect of the reconstructed object. poor. [0003] There are some related methods that can optimize the surface texture of static objects, but these methods are not suitable for dynamic objects. Some methods are able to reconstruct dynamic objects, but the reconstruction surface effect is poor. S...

AI Technical Summary

Problems solved by technology

The reconstruction of dynamic objects in related technologies pays more attention to the reconstruction of object motion, ignoring various errors caused by equipment, resulting in poor reconstructed surfaces.

Images