As most spectrum-based algorithms cannot effectively deal with the matching of size-variable point sets,a point pattern matching algorithm based on the Q-spectrum of line graph is proposed. In this algorithm,first,a weighted complete graph is constructed for each point set,and a line graph is constructed for each point by using the incident first k shortest edges. Then,a spectral decomposition is performed for the signless Laplacian matrix
constructed with the line graph,and the eigenvalues ( Q-spectrum) obtained from the spectral decomposition are used to represent the features of the point,which make it possible to calculated the matching probability. Finally,the optimal matching of point sets is searched by using the KM algorithm. Experimental results show that the proposed algorithm is of high matching accuracy,and that it can deal with the matching of two point sets with different sizes.