Abstract:Saliency detection has gained popularity in many applications, and many different approaches have been proposed. In this paper, we propose a new approach based on singular value decomposition (SVD) for saliency detection. Our algorithm considers both the human-perception mechanism and the relationship between the singular values of an image decomposed by SVD and its salient regions. The key concept of our proposed algorithms is based on the fact that salient regions are the important parts of an image. The singular values of an image are divided into three groups: large, intermediate, and small singular values. We propose the hypotheses that the large singular values mainly contain information about the non-salient background and slight information about the salient regions, while the intermediate singular values contain most or even all of the saliency information. The small singular values contain little or even none of the saliency information. These hypotheses are validated by experiments. By regularization based on the average information, regularization using the leading largest singular values or regularization based on machine learning, the salient regions will become more conspicuous. In our proposed approach, learning-based methods are proposed to improve the accuracy of detecting salient regions in images. Gaussian filters are also employed to enhance the saliency information. Experimental results prove that our methods based on SVD achieve superior performance compared to other state-of-the-art methods for human-eye fixations, as well as salient-object detection, in terms of the area under the receiver operating characteristic (ROC) curve (AUC) score, the linear correlation coefficient (CC) score, the normalized scan-path saliency (NSS) score, the F-measure score, and visual quality. (C) 2015 Elsevier Inc. All rights reserved.

Saliency analysis based on multi-scale wavelet decomposition

Efficient Saliency Analysis Based on Wavelet Transform and Entropy Theory

Saliency Analysis for Car Detection Based on 2-D Entropy and Velocity Prior

Visual Saliency via Multiscale Analysis in Frequency Domain and Its Applications to Ship Detection in Optical Satellite Images

Saliency Detection via Multi-Scale Global Cues

A multi-stage area saliency detection model

A Saliency Detection Model Using Low-Level Features Based on Wavelet Transform

Image Saliency Detection Based on Rectangular-Wave Spectrum Analysis

A Saliency Detection Method Based on Wavelet Transform and Simple Priors

Multi-scale Region-Based Saliency Detection Using W2 Distance on N-dimensional Normal Distributions.

Multi-scale saliency detection via inter-regional shortest colour path

Saliency Detection Using Wavelet Transform in Hypercomplex Domain

Saliency Region Detection Method Based on Background and Spatial Position

Saliency Detection Using Two-Stage Scoring

Visual Saliency Detection Based on Region Descriptors and Prior Knowledge

Multi-scale Hybrid Saliency Analysis for Region of Interest Detection in Very High Resolution Remote Sensing Images

Multi-Scale Salient-Region Detection for Optical Remote Sensing Images

Saliency detection using boundary information

Multiscale salient region detection and salient map generation for synthetic aperture radar image

Saliency Detection Based on Singular Value Decomposition

Saliency Detection Model Based on Wavelet Transform and Independent Component Analysis