Modeling deviations of rgb-d cameras for accurate depth map and color image registration
Multimedia Tools Appl.(2017)
摘要
A fundamental step to employ RGB-D cameras is to register color and depth images, whose misalignment may be caused by differences of camera poses and parameters, depth noises, etc. Previous methods mainly devote to more accurate camera calibration, which can only deal with misalignment that are parameterized with camera projection model. Other misalignment, which we call deviations , are more difficult to be measured and modeled. In this paper, we propose a method to model and remove RGB-D camera deviations. First, a specially-designed checkerboard with hollow squares is utilized to measure deviations and camera parameters, it takes advantage of the regularity of corner arrangements and can achieve high accuracy even with noisy depth inputs. Second, we propose a general deviation model to deal with irregular deviations that can not be handled by RGB-D camera projection model. Third, we introduce a registration method that incorporates the estimated deviation model to well register color and depth information. As demonstrated in the experiments, comparing with manufacturer’s calibration and some state-of-the-art algorithms, our approach produces significant better accuracy.
更多查看译文
关键词
RGB-D cameras, Camera calibration, Registration, Deviations, Computer vision
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络