The present paper is concerned with the improvement of existing and development of new approaches to the processing of experimentally obtained data by tracking individual particle. The approach can be applied in practice to address current scientific and applied problems in fluid mechanics and gas, energetics, transportation systems, etc. Each algorithm of the family of 3D PTV is divided into three parts: the search for the true position of particles in registered particle projections, the particle triangulation in 3D space and the matching of each particle in the first frame, a pair in the second frame. The study of three-dimensional flows has to use triangulation to find the three-dimensional coordinates of particles in a given volume by means of two or more projections with recording cameras. Modern three-dimensional experiments use 3 or 4 cameras to investigate flows in the majority of cases. To improve performance it is planned to develop new algorithms and modify an existing code to make use of high performance computing systems, particularly graphic processors. Also the improvement of the existing and creation of new algorithms for the reconstruction of three-dimensional distributions of particles in a flow volume was done. Besides the calculation of the velocity field was performed in order to reduce the resource intensity of the algorithms, to improve the accuracy and spatial resolution of the measured velocity field. Testing of the existing or created algorithms in this work, defining their characteristics, making recommendations for using of different approaches was carried out.