Automatic Image Segmentation with Superpixels and Image-Level Labels

In this work, we propose an automatic image segmentation algorithm based on super-pixels and image-level labels. Automatically and ideally segmenting the semantic region of each object in an image will greatly improve the precision and efficiency of subsequent image processing. We propose an automatic image segmentation algorithm based on super-pixels and image-level labels. The proposed algorithm consists of three stages. At the stage of super-pixel segmentation, we adaptively generate the initial number of super-pixels using the minimum spatial distance and the total number of pixels in the image. 

At the stage of super-pixel merging, we define small super-pixels and directly merge the most similar superpower pairs without considering the adjacency, until the number of super-pixels equals the number of groupings contained in image-level labels. Furthermore, we add a stage of reclassification of disconnected regions after super-pixel merging to enhance the connectivity of segmented regions. On the widely used Microsoft Research Cambridge data set and Berkeley segmentation data set, we demonstrate that our algorithm can produce high-precision image segmentation results compared with the state-of-the-art algorithms.