V.А. Kokovkina – Assistant,

P.G. Demidov Yaroslavl State University E-mail: thief_rus@yahoo.com

V.А. Antipov – Post-graduate Student,

P.G. Demidov Yaroslavl State University

E-mail: valant777@gmail.com

V.P. Kirnos – Senior Lecturer,

P.G. Demidov Yaroslavl State University

E-mail: crafter76@gmail.com

А.L. Priorov – Dr.Sc. (Eng.), Professor,

P.G. Demidov Yaroslavl State University

E-mail: andcat@yandex.ru

Е.D. Guryanov – Post-graduate Student,

P.G. Demidov Yaroslavl State University

E-mail: guryanoved@yandex.ru

One of the subtasks in the problem of simultaneous localization and mapping (SLAM) is described. The main attention is paid to algorithms for detecting landmarks based on contour analysis, whose work was studied as part of the software and hardware mobile complex.

Contour analysis is used to find landmarks. The idea of the approach is to refuse to process each point received from the laser rangefinder. The algorithms work with contours, each of which is assembled according to certain rules. The following algorithms based on contour analysis are considered: the method of matched filtering, the Teh-Chin algorithm, the Wu algorithm, and the curvature scale space method.

The method of matched filtering involves finding a quantitative measure of similarity between the filtered contour and a certain reference form. In this work, the reference form of the «corner» is used.

Other algorithms are aimed at finding the dominant points. Dominant points are points that have a high curvature. The Teh-Chin and Wu algorithms use a measure based on the cosine of the angle to calculate the curvature. Also, dominant points can be searched using the theory of the so-called scale space of curvature. This is the basis of the curvature scale space method.

Experimental studies were conducted in the virtual environment of Gazebo. In the simulated room one test race is held and the readings of the LIDAR are recorded. Then the recorded data is used to run the SLAM algorithm with the corresponding algorithm for detecting landmarks.

Initially, the algorithm parameters are adjusted, and at this stage the best results are shown by the Wu algorithm and the Teh-Chin algorithm.

The level of false-negative detection of landmarks, the rarer omission of landmarks, is lower for the matched filtering method and the curvature scale space method.

According to the accuracy of localization of the mobile object, the best results are shown by the method of matched filtering and the Wu algorithm. Also, they show the lowest sensitivity to noise (in particular, to the AWGN).

It is also worth noting that the algorithms for detecting landmarks based on contour analysis by the data of a LIDAR show accuracy at the level of visual methods for detecting landmarks, and in some cases even higher.

1. Gen K., Chulin N.A. Algoritm navigatsii bespilotnogo letatel'nogo apparata na osnove uluchshennogo algoritma odnovremennoy lokalizatsii i kartografirovaniya s adaptivnym lokal'nym diapazonom nablyudeniya. Vestnik MGTU im. N.E. Baumana. Ser. Priborostroenie. 2017. № 3. S. 76–94. [in Russian]
2. Dissanayake G., Durrant-Whyte H., Bailey T. Computationally efficient solution to the simultaneous localization and map building (SLAM) problem. Proceedings 2000 ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings, Australia, 2000.
3. Furman Ya.A., Krevetskiy A.V., Peredreev A.K., Rozhentsov A.A, Khafizov R.G., Egoshina I.L., Leukhin A.N. Vvedenie v konturnyy analiz. M.: 2003. S. 369–379. [in Russian]
4. Rattarengsi A., Chin R.T. Scale-Based Detection of corners of planar curves. IEEE Trans. Pattern Anal. Mach. Intell. 1992.
5. Ray B.K., Ray K.S. Detection of significant points and polygonal approximation of digitized curves. Pattern Recognition Lett. 13. 1992. P. 443–452.
6. Garrido A., Perez De La Blanca N., Garcia-Silvente M. Boundary simplification using a multiscale dominant-point detection algorithm. Pattern Recognition. 1998. V. 31. № 6. P. 791–804.
7. Teh C.H., Chin R.T. On the detection of dominant points on digital curves. IEEE Trans. Anal. Mach. Intell. 1989. P. 859–872.
8. Wu W.Y. An adaptive method for detecting dominant points. Pattern Recognition. 2003. V. 36. P. 2231–2237.
9. Witkin A.P. Scale-space filtering. Proc. 8th Int. Joint Conf. Artif. Intell., Karlsruhe, Germany. 1983. P. 1019–1022.
10. Li Y., Olson E. Extracting general-purpose features from LIDAR data. Proc. of IEEE Int’l Conf. on Robotics and Automation. Anchorage, AK, USA. May 2010.
11. Prozorov A.V., Priorov A.L. Trekhmernaya rekonstruktsiya stseny s primeneniem monokulyarnogo zreniya. Izmeritel'naya tekhnika. 2014. S. 24–28. [in Russian]
12. Prozorov A.V., Priorov A.L. Predobrabotka karty glubiny dlya povysheniya tochnosti pozitsionirovaniya kamery v zadache odnovremennoy lokalizatsii i kartirovaniya. Uspekhi sovremennoy radioelektroniki. 2016. № 4. S. 66–71. [in Russian]