一种多图像前景对应性的融合显著性检测算法研究

曲靖师范学院学报 ›› 2022, Vol. 41 ›› Issue (3): 75-81.

一种多图像前景对应性的融合显著性检测算法研究

刘磊

四川商务职业学院科研处,四川成都 610000

收稿日期:2022-03-18 出版日期:2022-05-26 发布日期:2022-06-02
作者简介:刘磊,四川商务职业学院科研处副教授,主要从事数字图像处理、机器学习研究.
基金资助:
四川高等职业教育研究中心课题“基于大数据的职业教育与乡村振兴协同发展研究”(GZY21B04).

A Joint Salience Detection Algorithm Based on Multi-image Foreground Correspondence

LIU Lei

Research Department, Sichuan Vocational College of Business, Chengdu Sichuan 610000,China

Received:2022-03-18 Published:2022-05-26 Online:2022-06-02

摘要/Abstract

摘要： 为过滤图像中的冗余信息,使显著性检测结果更加稳定,提出了一种面向多幅图像的基于前景对应性的融合显著性检测算法.首先对图像显著性区域进行粗略定位,然后采用边界连续性从凸包中选择准确的前景区域;其次在提取出前景显著图之后,对背景进行显著性检测;最后将对应前景区域与基于背景的显著图进行融合生成最后的显著图.结果显示,融合显著图的MAE值最低能够达到0.069,既优于其他算法,也优于单一检测的显著图的MAE值.说明结合背景的前景对应性显著性检测算法在多图像检测时具有良好的检测能力,在计算机视觉领域具有重要的研究意义和应用前景.

关键词: 多图像, 前景, 融合性显著, 显著性检测

Abstract: A joint salience detection algorithm based on foreground correspondence for multiple images is proposed to filter the redundant information in the image. The accurate foreground region is selected from the convex hull by using the boundary continuity, and then the corresponding foreground region is fused with the background based salience map to generate the final joint salience map. The results show that the minimum value of joint salience map can reach 0.071 because of other algorithms and the MAE value of single detection salience map, which indicates that the foreground correspondence salience detection algorithm combined with background has good detection ability in multi-image detection, and has important research significance and application prospect in the field of computer vision.

Key words: multi image, foreground, joint salience, salience detection

中图分类号:

TP391

刘磊. 一种多图像前景对应性的融合显著性检测算法研究[J]. 曲靖师范学院学报, 2022, 41(3): 75-81.

LIU Lei. A Joint Salience Detection Algorithm Based on Multi-image Foreground Correspondence[J]. Journal of Qujing Normal University, 2022, 41(3): 75-81.

参考文献

[1] 徐艺,凌赛广,董洲,等.一种基于计算机视觉的眼底图像质量评估系统的开发及应用[J].中华眼科杂志,2020,56(12):920-927.
[2] 李凤华,孙哲,牛犇,等.跨社交网络的隐私图片分享框架[J].通信学报,2019,40(7):1-13.
[3] 钟忺,陈纬航,钟珞.基于超像素及贝叶斯合并的图像分割算法[J].计算机工程与应用,2018,54(21):188-192+221.
[4] 王文冠,沈建冰,贾云得.视觉注意力检测综述[J].软件学报,2019,30(2):226-249.
[5] 李振强,王树才,赵世达,等.基于机器视觉和机器学习的羊骨架自动分割方法[J].食品与机械,2020,36(6):125-132.
[6] 刘根旺,张杰,张晰,等.基于残差谱显著性区域提取的极化SAR船只检测[J].电波科学学报,2019,34(6):751-760.
[7] TRAJANOVSKI S,SHAN C,WEIJTMANS P,et al.Tongue Tumor Detection in Hyperspectral Images Using Deep Learning Semantic Segmentation[J].IEEE Transactions on Biomedical Engineering,2021,68(4):1330-1340.
[8] 汪昊,刘向阳.基于区域中心的交互式图像前景提取方法[J].计算机技术与发展,2020,30(2):12-16.
[9] 仝苗,何坤,朱志娟.基于多尺度平滑的前景提取[J].四川大学学报(自然科学版),2020,57(2):271-276.
[10] 常振,段先华,鲁文超,等.基于多尺度的贝叶斯模型显著性检测[J].计算机工程与应用,2020,56(11):212-218.
[11] 陈宇,关可.结合前景先验和背景先验的显著性检测[J].电子设计工程,2020,28(9):154-157+166.
[12] 赵悉超,刘政怡,李炜.图像显著性传播及约束的协同显著性检测[J].计算机科学与探索,2019,13(11):130-139.
[13] 刘亚宁,吴清,魏雪.基于流行排序的前景背景显著性检测算法[J].科学技术与工程,2018,18(18):74-81.
[14] 张志龙,李爱华,李楚为.基于密度峰值搜索聚类的超像素分割算法[J].计算机学报,2020,43(1):1-15.
[15] WANG J,ZHOU J,WANG L,et al.Detection of residual yarn on spinning bobbins based on salient region detection[J].The Journal of the Textile Institute,2019,110(6):838-846.
[16] BF A,HC A.Context-aware long-term correlation tracking with hierarchical convolutional features[J].Pattern Recognition Letters,2019,127:110-118.
[17] KOSSÈ K E,ASSAF G J.Roads Funding Priority Index for Sub-Saharan Africa using Principal Components Analysis[J].Case Studies on Transport Policy,2019,7(4):732-748.
[18] 刘德鹏,李正周,曾靖杰,等.基于多尺度局部对比度和多尺度梯度一致性的红外小弱目标检测算法[J].兵工学报,2018,39(8):73-82.
[19] SINGH N.Saliency threshold: a novel saliency detection model using Ising's theory on Ferromagnetism(STIF)[J].Multimedia Systems,2020,26(6):397-411.
[20] CONG R,LEI J,FU H,et al.Going From RGB to RGBD Saliency:A Depth-Guided Transformation Model[J].IEEE Transactions on Cybernetics,2020,50(8):3627-3639.
[21] JIA N,ZHAO W,LIU X.Dempster hafer theory‐based hierarchical saliency detection[J].International Journal of Imaging Systems and Technology,2019,29(3):329-338.
[22] ZHOU Y,LI L,LING S,et al.Quality assessment for view synthesis using low-level and mid-level structural representation[J].Signal Processing Image Communication,2019,74:309-321.
[23] LI C,CHEN Z,WU Q,et al.Saliency object detection:integrating reconstruction and prior[J].Machine Vision and Applications,2019,30(3):397-406.
[24] LI X,YU J,DONG X,et al.Manifold ranking graph regularization non-negative matrix factorization with global and local structures[J].Pattern Analysis and Applications,2020,23(46):967-974.