摘要
图像显著性是实现视觉信息感知的最主要信息源,图像显著性信息的检测是计算机视觉研究中的热点.随着数据获取能力的极大提升,三维或立体视觉感知的需求越来越迫切,深度敏感信息的有效提取是影响当前立体视觉体验的重要元素,已经得到了很多研究人员的注意并取得了一定的研究成果.本文对图像显著性检测及深度敏感信息提取研究的相关检测模板及技术进行广泛研究,对深度敏感信息提取的研究现状进行了分类和概括,并通过对比实验,分析了几种典型深度敏感信息提取算法.最后,对深度敏感信息提取技术存在的问题及发展趋势进行了讨论.
Image saliency is the most important information source to realize visual information perception.The detection of image saliency information has always been a hot topic in computer vision research.With the great improvement of data acquisition capability,the demand of three-dimensional or stereoscopic vision is becoming more and more urgent.The effective extraction of deep sensitive information is an important element that affects the current stereoscopic vision experience.Therefore,it has been paid attention by many researchers and obtained some research achievements.In this paper,through extensive research on the relevant detection templates and technologies of image saliency detection and depth sensitive information extraction research,the current status of depth sensitive information extraction research is classified and summarized,and several typical depth sensitive information extraction algorithms are analyzed through comparative experiments.Finally,the problems and development trends of depth-sensitive information extraction techniques are discussed.
作者
牟海维
王妙妙
韩建
曹志民
Mu Haiwei;Wang Miaomiao;Han Jian;Cao Zhimin(School of Electronic Science,Northeast Petroleum University,Daqing,Heilongjiang 163318,China;Research and Development Center for Testing and Measurement Technology and Instrumentation,Heilongjiang University,Northeast Petroleum University,Daqing,Heilongjiang 163318,China)
出处
《激光与光电子学进展》
CSCD
北大核心
2020年第12期30-38,共9页
Laser & Optoelectronics Progress
基金
国家自然科学基金(51574087)
东北石油大学研究生创新科研项目(JYCX_CX09_2018)。
关键词
图像处理
深度敏感
显著性
立体视觉
视觉注意力
image processing
deep sensitivity
saliency
stereo vision
visual attention
