为解决三维场景重建问题，通过改进目前特征匹配的不足，提出了一种基于角仿射不变的特征匹配算法。
To solve the problem of 3-dimensional scene reconstruction, and to improve the performance of present feature point matching, a matching scheme which is invariant to perspective deformation induced...
最佳观测方位问题是计算机主动视觉研究的重要内容，广泛应用于计算机目标识别、摄影测量、三维场景重建等领域。
Next best view problem has become one of main influence factors in the vision research fields such as target recognition, stereo matching, visual tracking and scene reconstruction, etc.
利用车载激光扫描测量技术完成地形图修测、三维场景重建、建筑模型制作以及工程测量等任务，是近年来激光扫描测量技术应用研究的一个重要方向。
To complete Map Revision, 3d - Scene Rebuilding, Construction Modeling, Engineer Survey and other missions by using the vehicle laser - scanning technology is the important direction of research.
运用一个特定物体或者场景的多个图像可以减少图片的模糊度并且有可能更清晰地重建三维模型。
Using multiple images of a particular object or scene reduces ambiguity and makes possible more accurate three-dimensional reconstructions.
本文研究了基于图像的建筑物三维建模技术，主要分为：建筑物场景的全景图拼接与基于单幅图像的建筑物三维几何重建。
This paper has studied three dimensional modeling technology for architecture based on image, including panorama splicing and geometry reconstruction based on single image.
根据视觉导航应用的特点，提出了由视差图重建三维场景算法的误差分析及误差补偿算法。
From the specialties of application of vision-based navigation, we propose an algorithm for error analysis and compensate of reconstructing 3d scene from the disparity map.
根据立体视觉在视觉导航应用中的特点，提出了由视差图重建三维场景的整套算法。
From the specialties of the stereo vision applied in vision-based navigation, we propose a whole solution to reconstruct 3d scene from the disparity map.
我们实现的3D引擎主要包括三方面的内容：3D数据文件的读取，三维场景的重建，视点在场景中的运动。
The realization of our 3D engine is mainly consisted of three contents: reading 3D data file, rebuilding 3D scenes and eyes moving in the virtual scenes.
越来越多的应用如几何重建、碰撞检测、混合现实、手势识别等，都依赖于对三维场景准确且快速的分析。
An increasing number of applications depend on accurate and fast 3d scene analysis, such as geometry reconstruction, collision prevention, mixed reality, and gesture recognition.
越来越多的应用如几何重建、碰撞检测、混合现实、手势识别等，都依赖于对三维场景准确且快速的分析。
An increasing number of applications depend on accurate and fast 3d scene analysis, such as geometry reconstruction, collision prevention, mixed reality, and gesture recognition.
应用推荐