PROJECT TITLE :
Constrained Directed Graph Clustering and Segmentation Propagation for Multiple Foregrounds Cosegmentation
This paper proposes a brand new constrained directed graph clustering (DGC) method and segmentation propagation methodology for the multiple foreground cosegmentation. We have a tendency to solve the multiple object cosegmentation with the attitude of classification and propagation, where the classification is employed to obtain the item previous of each class and also the propagation is employed to propagate the previous to any or all pictures. In our methodology, the DGC technique is meant for the classification step, which adds clustering constraints in cosegmentation to stop the clustering of the noise knowledge. A new clustering criterion like the strongly connected component search on the graph is introduced. Moreover, a linear time strongly connected part search algorithm is proposed for the quick clustering performance. Then, we extract the article priors from the clusters, and propagate these priors to all the photographs to get the foreground maps, which are used to achieve the ultimate multiple objects extraction. We verify our method on each the cosegmentation and clustering tasks. The experimental results show that the proposed method can achieve larger accuracy compared with each the present cosegmentation methods and clustering ways.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here