Extraction method for image region of interest based on visual attention model

摘要

By means of the characteristics of the human visual system, this paper proposed a kind of extraction method for image region of interest (ROI) based on united global and local contrast. The significant calculation of the global contrast is executed through the optimized histogram contrast detection method. The significant calculation of the local contrast is executed through the local filtering by multiscale spatial Gaussian differential filter. Finally, the significant values are synthesized according to the basic requirements of the ROI. The experimental results show that our method gets a certain upgrade in extraction accuracy and operating efficiency for the ROI.