OBJECT DETECTION IN COMPUTER VISION SYSTEMS: A VISUAL SALIENCY BASED APPROACH

Abstract

A combined approach of object detection in image and eye fixation probability map calculation is proposed. This approach can be used in applied tasks of autonomous object detection. Experimental results show viability and efficiency of this approach as compared with state-of-art algorithms, and predict its usability on the broader class of tasks - applied variations of eye fixation problem.

Keywords

• визуальная салиентность
• центрально-периферический антагонизм
• эволюционная оптимизация
• задачи прогноза фиксации взгляда
• автономное обнаружение объектов