Guardado en:
Detalles Bibliográficos
Autores principales: Jin, Ji-Ping, Feng, Chen-Bin, Fan, Rui, Vong, Chi-Man
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2511.12084
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918212277370880
author Jin, Ji-Ping
Feng, Chen-Bin
Fan, Rui
Vong, Chi-Man
author_facet Jin, Ji-Ping
Feng, Chen-Bin
Fan, Rui
Vong, Chi-Man
contents Image stitching often faces challenges due to varying capture angles, positional differences, and object movements, leading to misalignments and visual discrepancies. Traditional seam carving methods neglect semantic information, causing disruptions in foreground continuity. We introduce SemanticStitch, a deep learning-based framework that incorporates semantic priors of foreground objects to preserve their integrity and enhance visual coherence. Our approach includes a novel loss function that emphasizes the semantic integrity of salient objects, significantly improving stitching quality. We also present two specialized real-world datasets to evaluate our method's effectiveness. Experimental results demonstrate substantial improvements over traditional techniques, providing robust support for practical applications.
format Preprint
id arxiv_https___arxiv_org_abs_2511_12084
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SemanticStitch: Enhancing Image Coherence through Foreground-Aware Seam Carving
Jin, Ji-Ping
Feng, Chen-Bin
Fan, Rui
Vong, Chi-Man
Computer Vision and Pattern Recognition
Image stitching often faces challenges due to varying capture angles, positional differences, and object movements, leading to misalignments and visual discrepancies. Traditional seam carving methods neglect semantic information, causing disruptions in foreground continuity. We introduce SemanticStitch, a deep learning-based framework that incorporates semantic priors of foreground objects to preserve their integrity and enhance visual coherence. Our approach includes a novel loss function that emphasizes the semantic integrity of salient objects, significantly improving stitching quality. We also present two specialized real-world datasets to evaluate our method's effectiveness. Experimental results demonstrate substantial improvements over traditional techniques, providing robust support for practical applications.
title SemanticStitch: Enhancing Image Coherence through Foreground-Aware Seam Carving
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.12084