Spatiotemporal Saliency Detection of Infrared Videos Based on Gestalt-guided Optimization

Authors

Wang Xin, 1. College of Computer and Information, Hohai University, Nanjing 211100, China;;2. Jiangsu Key Laboratory of Image and Video Understanding for Social Safety, Nanjing University of Science and Technology, Nanjing 210000, China;;
Chunyan Zhang, 1. College of Computer and Information, Hohai University, Nanjing 211100, China;;
Ning Chen, 3.School of Physics and Technology, Nanjing Normal University, Nanjing 210000, China;

Abstract

Abstract: A spatiotemporal saliency detection method based on Gestalt optimization is proposed. A method based on the multi-scale local sparse representation and local contrast measure is proposed to compute the spatial saliency in the infrared videos. A multi-frame symmetric difference approach is adopted to detect the temporal saliency. To get the initial spatiotemporal saliency map, a scheme based on the mutual-consistency is designed to fuse the spatial and temporal saliency maps adaptively. A Gestalt-guided optimization method is designed to calculate the final spatiotemporal saliency map. Experimental results show that the proposed method can effectively detect the spatiotemporal saliency of infrared videos.

Recommended Citation

Xin, Wang; Zhang, Chunyan; and Chen, Ning (2020) "Spatiotemporal Saliency Detection of Infrared Videos Based on Gestalt-guided Optimization," Journal of System Simulation: Vol. 32: Iss. 6, Article 5.
DOI: 10.16182/j.issn1004731x.joss.18-0668
Available at: https://dc-china-simulation.researchcommons.org/journal/vol32/iss6/5

First Page

1021

Revised Date

2018-12-19

DOI Link

https://doi.org/10.16182/j.issn1004731x.joss.18-0668

Last Page

1031

CLC

TP391.9

Recommended Citation

Wang Xin, Zhang Chunyan, Ning Chen. Spatiotemporal Saliency Detection of Infrared Videos Based on Gestalt-guided Optimization[J]. Journal of System Simulation, 2020, 32(6): 1021-1031.

DOI

10.16182/j.issn1004731x.joss.18-0668