Moving visual focus in salient object segmentation

Abstract

Saliency detection plays an important role in image segmentation, object detection and retrieval, which attracts more attention in the field of computer vision recently. Most existing saliency detection algorithms have not considered the influence of visual focus shifting yet. In this paper, a novel algorithm named moving region contrast (MRC) is proposed to analyze image saliency. The algorithm MRC is built on a novel concept of moving visual focus (MVF). The initial visual focus is defined as the geometric center of the image. Then the visual focus is calculated iteratively by focus-moving technique where a saliency gravitation model is employed to determine the moving direction. The salient region is obtained according to the final visual focus. The experiments are conducted on the dataset with 1000 images released by Achanta. Experimental results show that the proposed algorithm achieves marked improvements in performance and outperforms other 11 popular algorithms.

Robust saliency detection via corner information and an energy function

“…Top‐down [6–9] algorithms are task‐driven and require supervised learning with manually labelled ground truth. Bottom‐up methods [10–21] are mainly based on the low‐level visual information without any prior knowledge.…”