Region-division-based three-dimensional reconstruction method for texture image

Abstract

The invention discloses a region-division-based three-dimensional reconstruction method for a texture image. The method comprises the following steps: reading texture image information, performing region division on the texture image to obtain texture division regions, and performing region grouping; calculating height values of pixels in the texture image to obtain three-dimensional texture points of the pixels in the texture image by adopting a method of shape from shading; calculating the color roughness of each texture division region, determining the sizes of triangular meshes of different division regions, reconstructing a triangular patch of each division region, and converting the texture image into a three-dimensional texture graph to realize the three-dimensional reconstruction of the texture image. According to the method, the microscopic three-dimensional geometric shape of the surface of a virtual object reflected by the texture image is better reconstructed; microscopic details of the texture image are ensured, and complexity in the three-dimensional reconstruction of the texture image is further lowered; different physical attributes can further be endowed to different regions according to region grouping, so that a foundation is laid for the tactile sensation of the texture image.

Description

Technical field [0001] The invention relates to an image reconstruction method, in particular to a texture image three-dimensional reconstruction method based on region division, which can better reconstruct the microscopic three-dimensional geometric shape of the surface of a virtual object reflected by the texture image, and lays a foundation for the tactile perception of the texture image The foundation belongs to the field of computer graphics and virtual reality force tactile reproduction technology. Background technique [0002] Image information is mainly through human visual perception. However, for disadvantaged groups of blind people, it is difficult to see images through vision. Perception is the inevitable choice for blind people to perceive images. Blind people recognize images through perception functions such as touch or hearing instead of vision, and touch is the main alternative way of visual perception. Currently, texture image haptic rendering methods are most...

AI Technical Summary

Problems solved by technology

Currently, texture image tactile rendering methods are mostly based on 3D models, and 2D texture images lack height information, so they cannot be directly applied to tactile rendering methods

Images