计算机与数字工程2025，Vol.53Issue(1)：234-239,6.DOI:10.3969/j.issn.1672-9722.2025.01.042
基于语义一致性的精确目标移除
Precise Object Removal with Semantic Consistency
摘要
Abstract
Target removal is one of the hot topics of research in image inpainting with a wide range of applications.However,existing repair methods normally remove the object by drawing a rectangular bounding box or manually erasing on a pixel-by-pixel basis,which will cause additional damage to the background content.In this paper,a complete semantic segmentation-based target removal algorithm is proposed,which enables precise target pixel location by the semantic difference of image content before back-ground restoration,and achieves the goal of target removal while reducing the loss of background pixels.The experimental results show that the method can accurately remove the specific target from the image while obtaining semantically reasonable and textually plausible restoration results.关键词
图像修复/语义分割/目标移除Key words
image inpainting/semantic segmentation/object removal分类
信息技术与安全科学引用本文复制引用
王志宽,李昕..基于语义一致性的精确目标移除[J].计算机与数字工程,2025,53(1):234-239,6.基金项目
山东省自然科学基金项目(编号:ZR201911090059)资助. （编号:ZR201911090059）
