Locating the Wood Defects with Damped Newton Approaches based Non-negative Matrix Factorization

Non-negative matrix factorization (NMF) is an unsupervised method whose aim is to find an approximate factorization Vn*m=Wn*r*Hr*m into non-negative matrices Wn*r and Hr*m. This paper presents an extension to NMF and discusses the development and the use of damped Newton based the non-negative matrix factorization called DNNMF with good convergence property for wood defects detection. We also provide algorithms for computing these new factorizations and the supporting theoretical analysis. DNNMF can make sure the convergence of the cost function and has been tested with color wood images based on the statistical features extracted by local binary pattern (LBP). Finally, we present experimental results that explore the properties of DNNMF. Comparative experiments show DNNMF is effectual and practical with good research values and potential applications.