利索-全球专利检索

  1. 登录
  2. 注册
发明公开
CN106652044A 一种虚拟场景建模方法及系统 无效 - 驳回

一种虚拟场景建模方法及系统
摘要：
本申请公开了一种虚拟场景建模方法，包括：计算选取的至少两个待拼接图像的灰度值；根据所述待拼接图像的灰度值分别提取所述待拼接图像中灰度值相同的区域作为拼接区域；根据所述拼接区域对所述待拼接图像进行拼接；采用范数表示拼接后图像的梯度场能量函数，得到虚拟场景图像模型。针对现有技术中拼接算法的缺点，对其进行改进，采用范数来表示表示拼接后图像的图像梯度场能量函数，进而得到待拼接图像之间的梯度场差异，改进算法处理后的图像画面清晰，形成真实、清晰的虚拟现实场景模型。本申请还公开了一种虚拟场景建模系统，具有上述效果。
摘要（英）：

The invention provides a virtual scene modeling method. The method comprises the steps of calculating the gray-level values of at least two selected to-be-spliced images; according to the gray-level values of the to-be-spliced images, respectively extracting areas of the same gray-level value in the to-be-spliced images as splicing areas; according to the splicing areas, splicing the to-be-spliced images; and representing the gradient field energy function of the images after being spliced by adopting norms so as to obtain a virtual scene image model. According to the technical scheme of the invention, the defects of the existing splicing algorithm are overcome, so that the splicing algorithm is improved. The gradient field energy function of the images after being spliced are represented by adopting norms, and then the difference between the gradient fields of the images after being spliced is obtained. After the improved algorithm-based treatment, obtained images are clearer in frames, so that a true and clear virtual reality scene model is obtained. The invention also discloses a virtual scene modeling system having the effects described above.

0/0