Complex Network Construction for Interactive Image Segmentation using Particle Competition and Cooperation: A New Approach

In the interactive image segmentation task, the Particle Competition and\nCooperation (PCC) model is fed with a complex network, which is built from the\ninput image. In the network construction phase, a weight vector is needed to\ndefine the importance of each element in the feature set, which consists of\ncolor and location information of the corresponding pixels, thus demanding a\nspecialist's intervention. The present paper proposes the elimination of the\nweight vector through modifications in the network construction phase. The\nproposed model and the reference model, without the use of a weight vector,\nwere compared using 151 images extracted from the Grabcut dataset, the PASCAL\nVOC dataset and the Alpha matting dataset. Each model was applied 30 times to\neach image to obtain an error average. These simulations resulted in an error\nrate of only 0.49\\% when classifying pixels with the proposed model while the\nreference model had an error rate of 3.14\\%. The proposed method also presented\nless error variation in the diversity of the evaluated images, when compared to\nthe reference model.\n