TOP-DOWN VISUAL ATTENTION METHOD OF DETERMINING VISUAL ATTENTION REGION IN A STILL IMAGE

摘要

PURPOSE: A top-down visual attention method for determining a visual attention region in a still image is provided to increase successful probability of visual attention even when a major time feature of a target object is not identified. CONSTITUTION: A feature map and a stimulus map about a still image are determined(ST-2). The maximal stimulus area is determined from a stimulus map(ST-3). A feature map of a target object image is determined(ST-4). The feature map of the target object image is compared with the determined maximum stimulus area(ST-5). Visual features of high similarity are selected(ST-6). A final stimulus map is configured by combining stimulus maps of the visual features(ST-7). A view concentration area is determined from the stimulus map(ST-8).