Robust sparse representation based multi-focus image fusion with dictionary construction and local spatial consistency | Litlas