Image stitching strategy based on point-line feature fusion for homography estimation and seam-driven integration

Image stitching in complex scenes often fails due to the unconstrained nature of the environment. Recent studies have shown that seam-driven techniques can effectively eliminate artifacts in image stitching with large parallax. However, the effectiveness of seam-driven methods largely depends on the accuracy of the initial image alignment. To address this issue, this paper proposes a homography estimation method based on point-line feature fusion, incorporating line features to further enhance the alignment accuracy. After alignment, seam-driven techniques are employed as a post-processing strategy to eliminate residual artifacts in the stitching process. Experimental results demonstrate that this method not only significantly improves alignment accuracy but also effectively handles large parallax images, reducing artifacts caused by alignment errors.