An Effective Construction of Multiple Viewpoint based Panorama Design using Digital Image Processing Scheme
G. Ramkumar, T. J. Nagalakshmi, N. Nalini, P. Jagadeesh, P. Shyamala Bharathi, V. Amudha
Abstract
Panoramic stitching from a group of photos under wide baselines is the primary goal of this system, which uses SIFTS and SURF characteristics to achieve this goal. In this method, pictures and non-planar scene structures can have broad baselines. A mesh-based framework for 2D alignment and regularity is used instead of a 3D reconstruction. Composition should be improved, and the misaligned area should perform better. As a result, we have developed a new method called Novel image stitching approach for stitching together several photos collected from diverse perspectives. Our solution is unique in that it allows for large baselines between photos and non-planar scene structures. A mesh-based framework for 2D alignment and regularity is used instead of a 3D reconstruction. We may create panoramic pictures that are as perspective-preserving locally as feasible while yet being virtually orthogonal from a global perspective by solving an alignment-and-prior-constraint-based global objective function. On a misaligned area, we enhance composition and perform well. The proposed strategy has been proven to be successful in the face of difficult data sets.