基于去模糊和自适应边界HSV局部调整的纹理映射技术研究
发布时间：2019-02-18 16:47
【摘要】：文物数字化是计算机视觉在文化传承领域中非常重要的研究方向,其中三维重建和纹理映射是被应用得很广泛的技术,本文以实际应用场景为背景,针对当前纹理映射中存在的提升空间,提出了改善的方法,为文物数字化的发展提供更强有力的技术支持。本文首先对课题背景和现有技术进行了简要介绍;然后介绍了三维重建技术、纹理的表达模型、纹理的显示模型、纹理映射的发展情况;接下来阐述了单幅图像深度图重建方法、图像去模糊原理,并提出了一种利用多图像重建模型、原始图像和图像对应相机参数矩阵求解图像对应深度图,联合SML算子对图像进行分块的高斯反卷积实现全局去模糊的方法;论述了泊松方程的原理、RGB和HSV颜色模型,并提出了一种根据纹理块本身来自适应调整边界条带的宽度进行HSV颜色空间下V通道的泊松方程求解方法:最后介绍了应用本文所提的两种方法的纹理映射系统。从重建模型求解得到的深度图比一般方法得到的都要更加精准,去模糊的图像为最后的纹理映射带来了更好的结果,本文提出的修改过的局部颜色调整策略也对纹理映射的结果起到了正面作用。利用这一系统能让文物数字化的进程更加顺利。
[Abstract]:The digitization of cultural relics is a very important research direction of computer vision in the field of cultural inheritance, in which the three-dimensional reconstruction and texture mapping are widely used. In this paper, based on the actual application scene, this paper aims at the lifting space existing in the current texture mapping. The improved method is put forward to provide more powerful technical support for the digital development of the cultural relics. In this paper, the background of the subject and the existing technology are briefly introduced, then the three-dimensional reconstruction technique, the expression model of the texture, the display model of the texture and the development of the texture mapping are introduced. The reconstruction method of the single-image depth map and the principle of image de-blurring are described. and a method for realizing global deblurring by using a multi-image reconstruction model, an original image and an image corresponding camera parameter matrix to solve an image corresponding depth map and a joint SML operator to block an image are provided, and the principle, the RGB and the HSV color model of the Poisson equation are discussed, In this paper, a method for solving the Poisson equation of V-channel in HSV color space is presented based on the width of the texture block itself. The depth map obtained from the reconstruction model is more accurate than the general method, and the deblurred image brings better results for the final texture mapping. The modified local color adjustment strategy proposed in this paper also has a positive effect on the result of the texture mapping. With this system, the digital process of the cultural relics can be made more smooth.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41
本文编号：2426034
[Abstract]:The digitization of cultural relics is a very important research direction of computer vision in the field of cultural inheritance, in which the three-dimensional reconstruction and texture mapping are widely used. In this paper, based on the actual application scene, this paper aims at the lifting space existing in the current texture mapping. The improved method is put forward to provide more powerful technical support for the digital development of the cultural relics. In this paper, the background of the subject and the existing technology are briefly introduced, then the three-dimensional reconstruction technique, the expression model of the texture, the display model of the texture and the development of the texture mapping are introduced. The reconstruction method of the single-image depth map and the principle of image de-blurring are described. and a method for realizing global deblurring by using a multi-image reconstruction model, an original image and an image corresponding camera parameter matrix to solve an image corresponding depth map and a joint SML operator to block an image are provided, and the principle, the RGB and the HSV color model of the Poisson equation are discussed, In this paper, a method for solving the Poisson equation of V-channel in HSV color space is presented based on the width of the texture block itself. The depth map obtained from the reconstruction model is more accurate than the general method, and the deblurred image brings better results for the final texture mapping. The modified local color adjustment strategy proposed in this paper also has a positive effect on the result of the texture mapping. With this system, the digital process of the cultural relics can be made more smooth.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41
【参考文献】
相关期刊论文 前3条
1 王法强;耿国华;李康;贺毅岳;;基于特征点约束的人脸纹理映射[J];计算机应用研究;2012年09期
2 佟帅;徐晓刚;易成涛;邵承永;;基于视觉的三维重建技术综述[J];计算机应用研究;2011年07期
3 安新伟,张晓兵,尹涵春;医学图像三维重建的研究[J];电子器件;2001年03期
，本文编号：2426034
本文链接：https://www.wllwen.com/shoufeilunwen/xixikjs/2426034.html
