文献类型：期刊文献

英文题名：Oral endoscopic image stitching algorithm based on tooth region saliency perception

作者：Huang, Zhenyue[1]; Huang, Rong[1]; Chen, Jiayu[1]; Yan, Chunwei[1]; Chang, Qing[1]

机构：[1] School of Information Science and Engineering, East China University of Science and Technology, China

年份：2024

起止页码：11

外文期刊名：ACM International Conference Proceeding Series

收录：EI(收录号：20244317268084)

语种：英文

外文关键词：Diseases

摘要：Oral endoscope imaging is a highly simple, convenient, and cost-effective method for imaging the interior of the oral cavity. However, due to its hardware limitations, oral endoscopes often only provide localized imaging, which is not conducive to remote consultations and subsequent intelligent diagnostic assistance. The oral endoscope image stitching algorithm can obtain wide-field oral images through registration and stitching, meeting the needs of daily monitoring and diagnostic assistance. However, the handheld capture method results in significant visual disparities between images, leading to issues such as misalignment, ghosting, and the fragmentation of prominent objects in the stitched images, thereby affecting the diagnosis of related diseases. To address this issue, this paper proposes an optimal suture line search algorithm based on the saliency perception of tooth area regions. By extracting information from tooth area regions and designing relevant constraints, the optimal suture line is guided to pass through color and structurally similar areas while avoiding tooth areas as much as possible. This approach aims to better preserve the original image information and enhance the visual effect of the stitching. Experimental results demonstrate that the proposed optimal suture line stitching algorithm based on tooth area saliency perception better retains the detailed information of the original images, effectively improves the visual quality of the stitched images, and is more suitable for assisting in the diagnosis of oral diseases. ? 2024 Copyright held by the owner/author(s).