Detection, Recognition and Identification of Multiple Targets Dynamically with Multifoveated Images
Visual stimuli, multiresolution, multifoveation, similarity methods, interpolation
An image contains a large amount of visual information and when associated with a temporal sequence allows visualization of the dynamic existence of visual stimuli. However, several contexts of visual attention require that these stimuli be detected, recognized and/or identified. Given these contexts, a vision system with a single fovea needs to perform saccadic movements adding processing time to the task. In this work we propose a new approach to estimate the position of these visual stimuli employing the multifovation technique. For this, it is necessary to use the Multirresolution Multifoveation using Mobile Foveae method, features, interpolation techniques and similarity methods. We have found that a robotic system equipped with this approach will be better able to perform visual attention tasks in real time.