Probabilistic video stabilization using Kalman filtering and mosaicing

Litvin, Andrey; Konrad, Janusz; Karl, W.C.

doi:10.1117/12.476436

Cited by 154 publications

(113 citation statements)

References 7 publications

Supporting

Mentioning

111

Contrasting

Unclassified

Order By: Relevance

“…These methods can be classified into two types: one that uses the motion of objects on an image plane 7),10)-12), 18) and one that uses the orientation of a camera 15), 17) . Using the motion of objects on an image plane, Litvin, et al 10) and Matsushita, et al 11) have proposed methods that estimate the motion of textures in missing regions. Specifically, they use optical flows across the whole image and copy the pixel values to missing regions from different frames based on the estimated motion; however, it is difficult for these methods to determine appropriate optical flows in omnidirectional video because the appearance of textures changes between successive frames and the motion of pixels in large missing regions cannot be accurately estimated by such two-dimensional interpolation.…”

Section: Related Workmentioning

confidence: 99%

Video Completion for Generating Omnidirectional Video without Invisible Areas

Kawai

Machikita

Sato

et al. 2010

IPSJ Transactions on Computer Vision and Applications

View full text Add to dashboard Cite

Omnidirectional multi-camera systems cannot capture entire fields of view because of their inability to view areas directly below them. Such invisible areas in omnidirectional video decrease the resulting realistic sensation experienced when using a telepresence system. In this study, we generate omnidirectional video without invisible areas using an image completion technique. The proposed method compensates for the change in appearance of textures caused by camera motion and searches for appropriate exemplars considering threedimensional geometric information. In our experiments, the effectiveness of our proposed method has been demonstrated by successfully filling in missing regions in real video sequences captured using an omnidirectional multi-camera system.

show abstract

Section: Related Workmentioning

confidence: 99%

Video Completion for Generating Omnidirectional Video without Invisible Areas

Kawai

Machikita

Sato

et al. 2010

IPSJ Transactions on Computer Vision and Applications

View full text Add to dashboard Cite

show abstract

“…The goal of motion compensation is to remove high-frequency jitters from the estimated camera motion. It is the component that most video stabilization algorithms attempt to improve and many methods have been proposed, such as motion vector integration [1], particle filter [20], variational method [16], regularization [6], Kalman filter [11] and local parabolic fitting [8]. Image composition.…”

Section: Related Workmentioning

confidence: 99%

“…Most previous video stabilization methods focused on improving components of this framework. For example, many algorithms have been proposed for improving camera compensation by better smoothing out camera motion or recovering the intentional camera motion [1,6,11,16,20]. Some focused on improving image composition by motion inpainting and deblurring [14].…”

Section: Introductionmentioning

confidence: 99%

Video stabilization using robust feature trajectories

Lee

Chuang

Chen

et al. 2009

2009 IEEE 12th International Conference on Computer Vision

View full text Add to dashboard Cite

This paper proposes a new approach for video stabilization. Most existing video stabilization methods adopt a framework of three steps, motion estimation, motion compensation and image composition. Camera motion is often estimated based on pairwise registration between frames. Thus, these methods often assume static scenes or distant backgrounds. Furthermore, for scenes with moving objects, robust methods are required for finding the dominant motion. Such assumptions and judgements could lead to errors in motion parameters. Errors are compounded by motion compensation which smoothes motion parameters. This paper proposes a method to directly stabilize a video without explicitly estimating camera motion, thus assuming neither motion models nor dominant motion. The method first extracts robust feature trajectories from the input video. Optimization is then performed to find a set of transformations to smooth out these trajectories and stabilize the video. In addition, the optimization also considers quality of the stabilized video and selects a video with not only smooth camera motion but also less unfilled area after stabilization. Experiments show that our method can deal with complicated videos containing near, large and multiple moving objects.

show abstract

“…the intentional motion, and the undesired shaking or jitter, which are the only ones to be filtered to achieve a successfully video stabilization. Several techniques have been proposed to filter the shaking from the intentional motion such as Kalman filter [5] [7] and Motion Vector Integration [6]. However, they do not work properly when the intentional camera motion is fast and abrupt or when the magnitude of the camera shaking is variable along the time.…”

Section: Introductionmentioning

confidence: 99%

Automatic Feature-Based Stabilization of Video with Intentional Motion through a Particle Filter

del-Blanco

Jaureguizar

Salgado

et al. 2008

Advanced Concepts for Intelligent Vision Systems

View full text Add to dashboard Cite

Abstract. Video sequences acquired by a camera mounted on a hand held device or a mobile platform are affected by unwanted shakes and jitters. In this situation, the performance of video applications, such us motion segmentation and tracking, might dramatically be decreased. Several digital video stabilization approaches have been proposed to overcome this problem. However, they are mainly based on motion estimation techniques that are prone to errors, and thus affecting the stabilization performance. On the other hand, these techniques can only obtain a successfully stabilization if the intentional camera motion is smooth, since they incorrectly filter abrupt changes in the intentional motion. In this paper a novel video stabilization technique that overcomes the aforementioned problems is presented. The motion is estimated by means of a sophisticated featurebased technique that is robust to errors, which could bias the estimation. The unwanted camera motion is filtered, while the intentional motion is successfully preserved thanks to a Particle Filter framework that is able to deal with abrupt changes in the intentional motion. The obtained results confirm the effectiveness of the proposed algorithm.

show abstract

Probabilistic video stabilization using Kalman filtering and mosaicing

Cited by 154 publications

References 7 publications

Video Completion for Generating Omnidirectional Video without Invisible Areas

Video Completion for Generating Omnidirectional Video without Invisible Areas

Video stabilization using robust feature trajectories

Automatic Feature-Based Stabilization of Video with Intentional Motion through a Particle Filter

Contact Info

Product

Resources

About