Object Segmentation in Image Sequences Using Reinforcement Learning
Farhang Sahba · 2016
This paper proposes a reinforcement learning framework for the extraction of objects' borders in sequential images. In this method, an agent-based approach is proposed to find the appropriate local values and detect the edges of objects in the images. Due to its dynamic nature, this method could have potential for segmenting similar objects in image sequences that have the same nature.