基于人眼视觉规律的注视点分类及其在图像标注中的应用  

Visual-pattern-based fixation classification and its application in image annotation

作　　者：漆正溢 方红萍[1] 万中华 张瀚源 伍世虔 QI Zheng-yi;FANG Hong-ping;WAN Zhong-hua;ZHANG Han-yuan;WU Shi-qian(School of Information Science and Engineering,Wuhan University of Science and Technology,Wuhan 430081,China;School of Machinery and Automation,Wuhan University of Science and Technology,Wuhan 430081,China)

机构地区：[1]武汉科技大学信息科学与工程学院,湖北武汉430081 [2]武汉科技大学机械自动化学院,湖北武汉430081

出　　处：《液晶与显示》2023年第4期515-523,共9页Chinese Journal of Liquid Crystals and Displays

基　　金：国家自然科学基金(No.61775172);武汉科技大学重大科技项目培育计划(No.2018TDX06)。

摘　　要：针对现有眼动图像标注算法中停留在非目标上的注视点容易引入定位干扰,导致标注精度不高的问题,本文首先实验探索了标注任务中的眼动规律;然后提出将标注注视序列分为视觉搜索和视觉识别两个阶段,并设计了基于参数自适应DBSCAN算法的视觉搜索和视觉识别注视点分类方法,旨在将提取的识别注视点作为眼动图像标注算法的输入,提高标注结果的准确性;最后基于2014 DIMITRIOS P数据集开展实验对比与分析。实验结果表明,与现有相关算法相比,F1度量提升4%,算法运行效率提升了近1倍,眼动图像标注算法精度提高3.34%,满足稳定可靠、精度高、运行速度快等要求。In the existing works of eye-movement image annotation algorithms,the fixation points resting on non-targets may introduce localization interference,and produce low annotation accuracy.To solve this problem.Firstly,experimental studies are conducted to explore the eye-movement pattern in the annotation task.Then,the annotated gaze sequences are divided into two stages:visual search and visual recognition,and a fixation points classification method based on the parameter adaptive DBSCAN algorithm is proposed to extract recognition fixation points as the input of the eye-movement image annotation algorithm in order to improve the accuracy of the annotation results.Finally,the experimental comparison and analysis are carried out based on the 2014 DIMITRIOS P data set.The experimental results show that compared with the existing related algorithms,the F1 score is improved by 4%,the algorithm operation efficiency is improved by nearly one time,and the eye-movement image annotation accuracy is improved by 3.34%,which meets the requirements of stability,reliability,high accuracy,and fast running speed.

关 键 词：眼动注视点分类 眼动图像标注 视觉搜索 视觉识别 参数自适应DBSCAN 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]