Texturing is a process, in which surfaces of virtual models are covered with images (or fields) obtained either artificially, or by means of digital cameras. The process makes it possible to yield the effect of photorealism, which is especially desirable in numerous applications. The paper provides a review of methods of texturing, and discusses texturing-related issues: analysis of visibility, detection of blind spots. Special attention has been paid to the analysis of automatic methods as the most economic ones. The process itself generally makes use of collinearity equation in order to project triangles from a spatial polygon mesh onto oriented surfaces of images, performing visibility analysis. The relevant texture is assigned based on a comparative analysis of available textures. It is possible to texture a fragment of the model based on a single, optimal image or based on the "averaging" of many images. Research works were conducted in two aspects: searching for conditions of an optimal texturing, and testing programs for manual, semi-automatic and automatic texturing. Research carried out in a test field included experiments with the use of two methods of texturing, at variability of both texturized triangle sizes, and orientation of images made to provide textures. Following the research, it was possible to draw conclusions concerning the optimisation of texturing conditions. In the second part of research, different available programs were tested. The tests were conducted on exceptionally difficult objects, since the test data included 3D models of railway infrastructure objects and area close to the clearance envelope, as well as of the trackbed, obtained from mobile laser scanning data and associated digital colour images. The experiments have proved that presently there is no software available, which would enable an automatic generation of a coherent and correct mesh on the basis of point cloud generated by a laser scanner, and subsequently mapping images from a movie camera or a photo camera on the so-obtained mesh.