Boolean matrix factorization (BMF) is a data mining method that decomposes a binary matrix into the Boolean product of two smaller matrices, revealing hidden patterns in that matrix. Traditional BMF algorithms overlook visual aspects, limiting their ability to identify natural factors. This paper explores the use of image processing methods to enhance BMF. We propose treating Boolean matrices as black-and-white images and applying methods such as analysis connected component, SLIC segmentation, Quad tree, spatial filtering, and morphological operations. Experiments on real and synthetic datasets demonstrate that these methods highlight relevant structures in data and improves factor computation.