Nowadays X-ray computed tomography is widely used in medicine for solving various diagnostic problems. At the same time there is a large group of software whose function is to automate diagnostic processes, also known as computer-aided diagnosis software. However, there are still some processes which have not been automated yet. This concerns the parameters for calculating small objects of examination, such as pathological tumors, sinuses and so on, which occupy only a small part of the tomographic image. The known methods of automated image processing do not allow carrying out identification and parametrization of such objects. This study has proposed the algorithm and its software implementation allowing to visualize small objects in tomographic images, detect the boundaries of these objects and calculate their parameters with a high precision. This data can help medical professionals to make the diagnosis more precisely and faster and to reduce the number of repeated tomographic tests.