Back close

Multi-resolution dynamic mode decomposition-based salient region detection in noisy images

Publication Type : Journal Article

Publisher : Signal, Image and Video Processing

Source : Signal, Image and Video Processing, Springer London (2020)

Url : https://www2.scopus.com/inward/record.uri?eid=2-s2.0-85069851519&doi=10.1007%2fs11760-019-01539-9&partnerID=40&md5=458828e65f540513bc40dad3b9cb9cb6

Keywords : Adaptive image compressions, Behavioral research, Computational model, Computer vision applications, Dynamic mode decompositions, Image compression, Noisy image, Object Detection, Object recognition, Salient object detection, Salient region detections, Visual Attention

Campus : Coimbatore

School : Department of Computer Science and Engineering, School of Engineering

Center : Center for Computational Engineering and Networking, Computational Engineering and Networking

Department : Electronics and Communication

Year : 2020

Abstract : Detection of salient region in an image is a crucial problem in many cognition and computer vision applications like object detection, adaptive image compression, automatic image cropping, video and image analysis. A part of an image is considered as salient, if the set of pixels under consideration protrudes from the rest, in terms of features such as color, contrast and local orientations. Generally, computational models for saliency assume that the image under observation is clean and fails to deal with visual disturbances. This paper presents a robust method for the detection of salient regions, using the multi-resolution dynamic mode decomposition (MRDMD approach). Effectiveness of the proffered method for the detection of salient region within clean and noisy images was examined and successfully verified for a wide range of noise strengths. © 2019, Springer-Verlag London Ltd., part of Springer Nature.

Notes: cited By 0

Cite this Research Publication : O. K. Sikha and Dr. Soman K. P., “Multi-resolution Dynamic Mode Decomposition-based Salient Region Detection in Noisy Images”, Signal, Image and Video Processing, 2020.

Admissions Apply Now