Color Correction for an Image Sequence
IEEE Computer Graphics and Applications
Color Histogram Correction for Panoramic Images
VSMM '01 Proceedings of the Seventh International Conference on Virtual Systems and Multimedia (VSMM'01)
Automatic Panoramic Image Stitching using Invariant Features
International Journal of Computer Vision
Outdoors augmented reality on mobile phone using loxel-based visual feature organization
MIR '08 Proceedings of the 1st ACM international conference on Multimedia information retrieval
Panorama Mosaic Optimization for Mobile Camera Systems
IEEE Transactions on Consumer Electronics
ACM SIGGRAPH ASIA 2009 Posters
Fast and high-quality image blending on mobile phones
CCNC'10 Proceedings of the 7th IEEE conference on Consumer communications and networking conference
Color and luminance compensation for mobile panorama construction
Proceedings of the international conference on Multimedia
Bagadus: An integrated real-time system for soccer analytics
ACM Transactions on Multimedia Computing, Communications, and Applications (TOMCCAP) - Special issue of best papers of ACM MMSys 2013 and ACM NOSSDAV 2013
Hi-index | 0.00 |
This paper presents a simple and efficient color and luminance compensation approach for image sequences for constructing panoramic images on mobile devices. In this approach, constant compensation coefficients for adjacent images are computed from the corresponding pixels in the overlapping areas of the adjacent images in the linearized RGB color space. A global adjustment for all compensation coefficients is performed so that the correction for each image is as small as possible. This can smooth color transition between adjacent images in the image sequence globally and reduce cumulative errors in the color correction process. The compensation coefficients, together with the global adjustment factor, are applied to correct color and luminance of source images in the image sequence before they are stitched into a panoramic image. Our algorithm is not sensitive to the quality of spatial registration since it does not need exact pixel correspondences between the overlapped images. The computation of compensation coefficients for each adjacent image takes the gamma correction factor into account, producing better results. Running the method separately on the three channels matches the color balance that may not match due to auto-white-balancing algorithm. A global adjustment for the compensation coefficients minimizes color transitions and reduces cumulative errors. The approach is integrated into a sequential image stitching procedure and implemented in a mobile panorama system, good results have been obtained for both indoor and outdoor scenes.