立体显示时，视差图像中过大的竖直视差将对人眼立体视觉造成干扰，容易使观看者产生视觉疲劳。
In three-dimensional displays, large vertical parallax in parallax images is disadvantageous to stereo vision due to the presence of visual fatigue.
基于各视差图像中同源像点对应同一物点的原理，本文提出了一种消除多视点视差图像竖直视差的方法。
Based on the principle that homologous points in different parallax images correspond to the same object point, a method is proposed to eliminate the vertical parallax in multi-view parallax images.
本文从人类立体视觉的原理出发，重点讨论在计算机中如何从3d模型获取用于裸眼立体显示视差图像的问题。
The problem of how to extract parallax images from computer 3d models used in stereoscopic display is discussed in this paper.
为了解决视差图像的快速、鲁棒配准问题，提出了一种基于小波近似图像和垂直细节图像由粗到精的加权全局配准策略。
To solve the quick and robust registration of parallax images, a coarse-to-fine global registration strategy using the coarse images weighted with detail image was proposed.
针对双镜立体照相技术重构立体图像画质模糊、图像受损、清晰度下降的问题，采用光开关对镜头光路左右对称分割，获得视差图像。
For solving the problem of ghost image recurred on 3DTV, the light-path of lens is symmetrically divided by optical switch to obtain parallax image.
而后用图像通过滤波器来观察，即所谓的视差屏障（parallaxbarrier），其实主要就是一个划有很多垂直狭缝的观察屏。
The images are then viewed through a kind of filter--called the parallax barrier--that essentially is a screen with a bunch of vertical slits in it.
所以应该明确，以避免视差，这是重要的是要考虑所有的图像从同一观点。但有一个例外。
So it should be clear that in order to avoid parallax, it is important to take all images from the same viewpoint. But there is one exception.
本文提出一种新的用于立体图像编码的视差估计和遮挡点检测混合算法。
This paper proposes a novel cooperative algorithm for disparity estimation and occlusion points detection, which is adaptive to stereo image coding.
这种技术利用所谓的“视差栅栏”，使两只眼睛分别接受不同的图像，来形成立体效果。
The technology adopts the so-called "parallax barrier" to make two eyes respectively receive different images to achieve the three-dimensional effect.
视差滚动使背景图像比前景的内容移动得更慢，继而产生深度和沉浸的错觉。
Parallax scrolling involves the background image moving slower than the content in the foreground, creating an illusion of depth and immersion. Image credits: squarespace.
在能够求得最佳窗口和正确视差的前提下，该方法多次利用整数图像来提高效率，使得计算量与窗口大小无关。
On the premise of a sure find of an optimal window size, The method utilizes integral image to speed up matching computation, so the calculation is independent of the window size.
并在分析了视差和残差图像的统计特性后，提出了一个基于小波变换的视差补偿残差(WDCD)的编码方案。
After analyzed the statistical characteristics of disparity and residual image, a new coding scheme of disparity compensated difference based on wavelet transform (WDCD) is presented.
视差估计与补偿是立体图像编码的关键问题，本文提出了一种基于交叠块匹配的视差估计与补偿算法。
Disparity estimation and compensation are vital tasks in stereo image coding. An algorithm for disparity estimation and compensation based on overlapped block matching is proposed in this paper.
介绍了基于视差理论的线条光栅并像、分像作用和远、近景物的立体成像技术，以及立体图像的制作方法。
Depends on disparity theory, it is given the function of combining and separating image with line grating, stereo imaging of far, near objects and its making method.
针对传统方法难以可靠估计图像中纹理单一像素点视差的问题，提出一种新的基于纹理分析的视差估计算法。
According to the problem that it is difficult for traditional algorithms to get disparity of less textured pixels, a novel stereo matching algorithm based on texture analysis is proposed.
核心技术有视差估计、遮挡检测、立体残差图像编码、运动和视差的联合估计等。
The key techniques include disparity 'estimation, occlusion detection, disparity compensated difference coding and joint estimation of disparity and motion.
然后根据图像和视差范围的大小分层，建立由粗及细的自适应分层图像金字塔结构，加快搜索速度、减少错误匹配；
Then, according to restriction of the image size and the disparity range, the coarse to fine adaptive hierarchical image pyramids have been built to search fast and reduce wrong matching.
并针对图像的深度连续性约束，定义了一个反映图像自身特征的新能量函数，应用于图割算法，提高了视差计算精度。
Moreover, to improve the disparity accuracy, a novel energy function is well designed to restrict the depth-continuity of an image and applied to the Graph Cut algorithm to describe the image feature.
实验证明，这种方法可以通过双焦立体图像对高效地获得场景的视差图。
The experimental results show that this method can get disparity map efficiently from pairs of bi-focal images.
首先根据尺度不变特征变换方法从图像中提取关键点作为特征点，然后进行左右双目图像的特征点匹配和视差的恢复。
Then use these feature points for binocular matching thus getting distance information of the scene by the calculation of differences between match points in two images.
首先根据尺度不变特征变换方法从图像中提取关键点作为特征点，然后进行左右双目图像的特征点匹配和视差的恢复。
Then use these feature points for binocular matching thus getting distance information of the scene by the calculation of differences between match points in two images.
应用推荐