True Multi-Image Alignment and Its Application to Mosaicing and Lens Distortion Correction
IEEE Transactions on Pattern Analysis and Machine Intelligence
Underwater video mosaics as visual navigation maps
Computer Vision and Image Understanding - Special issue on underwater computer vision and pattern recognition
Video orbits of the projective group a simple approach to featureless estimation of parameters
IEEE Transactions on Image Processing
Efficient, robust, and fast global motion estimation for video coding
IEEE Transactions on Image Processing
A novel CBCD scheme based on local features category
WISM'12 Proceedings of the 2012 international conference on Web Information Systems and Mining
| Hi-index | 0.00 |
This paper presents a double hierarchical algorithm to construct large view mosaics of videos, taken with a camera that is free to pan, tilt, rotate, and zoom. Our method consists of two different hierarchical procedures. The first one deals with the estimation of global motion parameters between successive pairs of images. This is a modification of Mann's method based on a linear hierarchical technique combined with the associative law in group theory. The second one registers each three consequent images into an intermediate submosaic for the next hierarchical level. In order to improve the precision of the coordinate transformations between two frames we develop a local adjustment technique and a global feedback technique. The techniques significantly reduces the number of compositions of coordinate transformations from each frame to the reference frame, which results in significant decrease of accumulated registration errors mainly caused by rounding off in each matrix product. Plentiful experimental results show that our double hierarchical algorithm can efficiently control accumulated registration errors, thereby allowing the creation of high quality panoramic mosaics for perspective videos with great change of view.