The Idea of Non-negative Matrix Factorization (NMF) has been implemented in a wide variety of real world applications. To improve the usability of NMF, people usually add some regularization items to constrain the process of matrix factorization. The Regularized Non-negative Matrix Factorization (RNMF) mainly relies on prior knowledge to set the regularization parameter value. An improper regularization parameter value will directly influence the factorization result. Due to this reason, we propose a novel algorithm based on L-curve theory which is able to determine the regularization parameter value automatically, we name the algorithm as autoNMF. To testify the validity of autoNMF, we experiment the algorithm with the synthetic image data in blind source separation and compare it with other algorithms. Better unmixed results are gained indicating our algorithm outperforms other algorithms in several aspects.