Wyniki wyszukiwania - BazTech

1

A set of depth sensor processing ROS tools for wheeled mobile robot navigation

Drwięga M., Jakubiak J.

Journal of Automation Mobile Robotics and Intelligent Systems

|

2017

|

Vol. 11, No. 2

48--56

EN

The paper presents a set of soŌware tools dedicated to support mobile robot navigaƟon. The tools are used to process an image from a depth sensor. They are implemented in ROS framework and they are compaƟble with standard ROS navigaƟon packages. The soŌware is released with an open source licence. First of the tools converts a 3D depth image to a 2D scan in polar coordinates. It provides projecƟon of the obstacles, removes the ground plane from the image and compensates sensor Ɵlt angle. The node is faster than the standard node within ROS and it has addiƟonal funcƟons increasing range of possible applicaƟons. The second tool allows detecƟon of negaƟve obstacles i.e. located below the ground plane level. The third tool esƟmates height and orientaƟon of the sensor with RANSAC algorithm applied to the depth image. The paper presents also the results of usage of the tools with mobile plaƞorms equipped with MicrosoŌ Kinect sensors. The plaƞorms are elements of the ReMeDi project within which the soŌware was developed.

2

Zestaw narzędzi wspomagających nawigację kołowym robotem mobilnym wyposażonym w sensor głębi

Drwięga M., Jakubiak J.

Prace Naukowe Politechniki Warszawskiej. Elektronika

|

2016

|

z. 195, t. 2

437--446

PL

Niniejszy referat prezentuje zestaw narzędzi, które powstały w celu wspomagania systemu autonomicznej nawigacji kołowej platformy mobilnej wyposażonej w sensor głębi. Dostępne funkcjonalności dotyczą między innymi konwersji obrazu głębi do postaci dwuwymiarowej we współrzędnych biegunowych, przy czym z mapy głębi usuwane jest podłoże i przeprowadzana jest kompensacja wpływu pochylenia czujnika na zwracane odległości. Dodatkowo, zestaw zawiera narzędzie do wykrywania przeszkód wklęsłych jak schody czy urwiska. Ostatni pakiet oprogramowania służy do estymacji wysokości oraz kąta pochylenia czujnika głębi względem podłoża na podstawie obrazu głębi. Narzędzia zaimplementowano w środowisku ROS i są kompatybilne z czujnikiem Microsoft Kinect.

EN

The paper presents set of tools based depth sensor for navigation system of ReMeDi mobile platform. First of the tools allows to convert a 3D depth image to a 2D scan in polar coordinates, to remove ground plane from the image and to compensate sensor tilt angle. Moreover, paper presents a method of negative obstacles detection based on depth sensor. The method is compatible with standard ROS navigation package. The last tool is used for the depth sensor pose estimation with respect to the ground using the RANSAC algorithm. The tools were implemented in ROS environment and they were tested with Microsoft Kinect sensor.

3

Significance of features in object recognition using depth sensors

Harasymowicz-Boggio B., Chechlinski L, Siemiatkowska B.

Optica Applicata

|

2015

|

Vol. 45, nr 4

559--571

EN

This article concerns a key topic in the field of visual object recognition – the use of features. Object recognition algorithms typically rely on a fixed vector of pre-selected features extracted from 2D or 3D scenes, which are then analyzed with various classification techniques. On the other hand, the activation of particular features in biological vision systems is hierarchical and data-driven. To achieve a deeper understanding of the subject, we have introduced several mathematical tools to estimate multiple RGB-D features’ relevance for different object recognition tasks and conducted statistical experiments involving our database of high quality 3D point clouds. From the thorough analysis of the obtained results we draw conclusions that may be useful to design better, more adaptive object recognition algorithms.