A novel approach to combine features for salient object detection using constrained particle swarm optimization

作者：

Highlights：

• We have used constrained particle swarm optimization approach to combine features.

• A new objective function is proposed to compute the weights.

• An adaptive threshold is used for pixel classification instead of fixed threshold.

• Our model has the best precision, recall, F -measure and AUC values.

摘要

•We have used constrained particle swarm optimization approach to combine features.•A new objective function is proposed to compute the weights.•An adaptive threshold is used for pixel classification instead of fixed threshold.•Our model has the best precision, recall, F -measure and AUC values.

论文关键词：Salient object detection,Particle swarm optimization,Multi-objective function

论文评审过程：Received 4 April 2013, Revised 11 September 2013, Accepted 7 November 2013, Available online 20 November 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.11.012