A New Saliency Object Extraction Algorithm Based on Itti’s Model and Region Growing
Yunwei Jia, Chenxiang Hao, Kun Wang · 2019
A new method based on Itti's model is proposed in order to extract saliency objects as complete as possible. It combines the advantages of Itti's model and region growing. Firstly, visual features of early such as intensity, color and orientation are extracted from the input static color image. Secondly, three conspicuity maps are created according to early features. Thirdly, in order to obtain the initial seed points, thresholds based on three conspicuity maps are set respectively. Fourthly, three maps after region growing are attained. Finally, three maps which are based on region growing, are combined into a saliency map. A series of experiments has demonstrated that the algorithm is effective. Compared with Itti's model, the precision, the recall rate and F-measure of the saliency object extraction by the proposed algorithm are improved obviously.