Scientific and Technical Aspects of Applying the Method of Principal Components in the Processing of Lidar Data
DOI:
https://doi.org/10.22213/2410-9304-2018-1-4-10Keywords:
laser scanning, time-of-flight camera, cluster analysis data, treatment parameters, principal component analysis, route mapAbstract
The processing of experimental data for the purpose of subsequent decision making in the field of mobile robot control is one of the complex processes that require the use of modern algorithms and mathematical methods. Most often, the data on mobile robot positioning are formed using machine vision systems. One of the problems in managing mobile robots is the timely determination of the coordinates of the mobile robot's location and obstacles to its movement. The considered system of technical vision based on LIDAR realizes the spatial orientation of the mobile robot and forms a single technical module. This module has a number of advantages in processing experimental data of industrial premises that have a significant area and are represented by a number of dynamic and static objects. The paper proposed the use of agglomerative hierarchical clustering algorithm using the method of principal components for processing the position data of the robot, obtained by means of a laser rangefinder Hokuyo UTM-30LN.The paper is devoted to the development of a scanning system and the subsequent formation of a route map, using two parameters of the technical vision system - range and intensity. The paper contains a mathematical description of the clustering algorithm and implemented the process of constructing a route map of the mobile robot using the data obtained with the LIDAR-system. The results of real data processing are shown, which prove the effectiveness of the modified algorithm using LIDAR-parameters: the range to the object and light intensity.References
Sergi F., Guillem A., Juan Andrade-Cetto, Carme T. Object modeling using a ToF camera under an uncertainty reduction approach // IEEE International Conferenceon Roboticsand Automation (ICRA). IEEE, 2010. C. 1306-1312.
Абрамов И. В., Абрамов А. И., Мазитов Т. М., Пальмов А. М. Применение пчелиного алгоритма для обработки данных лазерной сканирующей системы при навигации мобильных роботов // Вестник ИжГТУ имени М. Т. Калашникова. 2016. Т. 19. № 2. С. 101-104.
Абрамов И. В., Абрамов А. И., Мазитов Т. М. Модификация алгоритма ICP путем внедрения коэффициента усиления для ускорения совмещения облаков точек // Интеллектуальные системы в производстве. 2016. Т.14. № 2. С. 4-9.
Баранова И. В., Ооржак О. Е. Алгоритм кластеризации k-средних и его реализация в среде MATLAB // Сборник научных трудов XI Международной конференции студентов и молодых ученых. Томск, 2014. С. 645-647.
Huang Zhexue. Extensions to the k-means algorithm for clustering large data sets with categorical values // Data Mining and Knowledge Discovery. 1998. Т. 2. № 3. С. 283-304.
Ralambondrainy H. A conceptual version of the K-Means algorithm // Pattern Recognition Letters. 1995. Т. 16, № 11. С. 1147-1157.
San O. M., Huynh V. N., Nakamori Y. An alternative extension of the k-means algorithm for clustering categorical data // International JournalApplied Mathematic Computing Science. 2004. Т. 14, № 2. С. 241-247.
Jolliffe, I. T. Principal Component Analysis. Series: Springer Series in Statistics. М. : Springer, NY, 2002. 487 с.
Сканирующий лазерный дальномер UTM-30LN. URL: http://sensotek.ru/images/hokuyo/ UTM-30LX_spec_en.pdf (дата обращения: 04.11.2017).
Michael Steinbach, George Karypis, Vipin Kumar. A Comparison of Document Clustering Techniques. Department of Computer Science and Egineering. University of Minnesota, 2000. 20 c.
Воронцов К. В. Лекции по алгоритмам кластеризации и многомерного шкалирования. URL: http://www.machinelearning.ru/wiki/images/ c/ca/Voron-ML-Clustering.pdf (дата обращения: 14.11.2017).
Кострикин А. И., Манин Ю. И. Линейная алгебра и геометрия. 5-е изд., перераб. и доп. М. : Лань, 2008. 304 с.
