Peter Beňo

This paper is focused on the proposal and verification of the RGBD mapping system for a small, low-cost mobile robot. The solution's requested properties were easy to replicate and easy to extend for further development on commonly available personal computers. The proposed solution is based on a Kinect sensor. Furthermore, 14 feature detectors were evaluated, and an ORB detector was chosen for the final implementation. In the image, pre-processing CLAHE filter was applied. Post-processing used… Show more

This paper is focused on the proposal and verification of the RGBD mapping system for a small, low-cost mobile robot. The solution's requested properties were easy to replicate and easy to extend for further development on commonly available personal computers. The proposed solution is based on a Kinect sensor. Furthermore, 14 feature detectors were evaluated, and an ORB detector was chosen for the final implementation. In the image, pre-processing CLAHE filter was applied. Post-processing used the modification of the RANSAC method. The final solution proves a globally consistent SLAM based on an RGBD sensor. The article also presents research, which suggests a parallelization scheme of the computational process using a multi-core CPU to achieve real-time processing. Show less

This paper presents novel extensions of the Stochastic Optimization Motion Planning (STOMP), which considers cartesian path constraints. It potentially has high usage in many autonomous applications with robotic arms, where preservation or minimization of tool-point rotation is required. The original STOMP algorithm is unable to use the cartesian path constraints in a trajectory generation because it works only in robot joint space. Therefore, the designed solution, described in this paper… Show more

This paper presents novel extensions of the Stochastic Optimization Motion Planning (STOMP), which considers cartesian path constraints. It potentially has high usage in many autonomous applications with robotic arms, where preservation or minimization of tool-point rotation is required. The original STOMP algorithm is unable to use the cartesian path constraints in a trajectory generation because it works only in robot joint space. Therefore, the designed solution, described in this paper, extends the most important parts of the algorithm to take into account cartesian constraints. The new sampling noise generator generates trajectory samples in cartesian space, while the new cost function evaluates them and minimizes traversed distance and rotation change of the tool-point in the resulting trajectory. These improvements are verified with simple experiments and the solution is compared with the original STOMP. Results of the experiments show that the implementation satisfies the cartesian constraints requirements. Show less

The objective of this article is to propose and verify a reliable detection mechanism of multi-rotor unmanned aerial vehicles (UAVs). Such a task needs to be solved in many areas such as in the protection of vulnerable buildings or in the protection of privacy. Our system was firstly realized by standard computer vision methods using the Oriented FAST and Rotated BRIEF (ORB) feature detector. Due to the low success rate achieved in real-world conditions, the machine-learning approach was used… Show more

The objective of this article is to propose and verify a reliable detection mechanism of multi-rotor unmanned aerial vehicles (UAVs). Such a task needs to be solved in many areas such as in the protection of vulnerable buildings or in the protection of privacy. Our system was firstly realized by standard computer vision methods using the Oriented FAST and Rotated BRIEF (ORB) feature detector. Due to the low success rate achieved in real-world conditions, the machine-learning approach was used as an alternative detection mechanism. The “Common Objects in Context dataset” was used as a predefined dataset and it was extended by 1000 samples of UAVs from the SafeShore dataset. The effectiveness and the reliability of our system are proven by four basic experiments—drone in a static image and videos which are displaying a drone in the sky, multiple drones in one image, and a drone with another flying object in the sky. The successful detection rate achieved was 97.3% in optimal conditions. Show less

The main idea of proposed paper is to specify and verify the segmentation method (or a group of segmentation techniques) ideal for tracking a moving object in a scene. The object is represented by an animal (guinea pig) in a laboratory cage. Object tracking based on visual system uses and compares different techniques of segmentation: global thresholding, adaptive thresholding, differential operators used for motion detection in a video sequence and color matching algorithm implemented from… Show more

The main idea of proposed paper is to specify and verify the segmentation method (or a group of segmentation techniques) ideal for tracking a moving object in a scene. The object is represented by an animal (guinea pig) in a laboratory cage. Object tracking based on visual system uses and compares different techniques of segmentation: global thresholding, adaptive thresholding, differential operators used for motion detection in a video sequence and color matching algorithm implemented from industrial applications. Animal activity is a significant marker corresponding with health status. Trajectory of an animal is a good quantitative parameter for evaluation of its activity. In our case, animal tracking was used for induced gastrointestinal diseases (gastric ulcer) in the group of guinea pigs. The results serve as a supporting tool in pharmacology and pathological physiology to set optimal therapy and medication for selected groups of diseases and also help to understand the mechanisms of diseases in the medical research. Experimental measurements were taken in the large database of healthy (reference) animals and “non-healthy” ones. The visual system-based animal (object) tracking could be used as a non-expensive and flexible (reconfigurable) replacement of radio-frequency identifiers. The paper also discusses the possibility to use introduced methods in real-time applications.

Show less

Abstract This article deals with tracking of laboratory animals (guinea pigs] in cage. Authors
collective applied machine vision tools based on global automatic thresholding, adaptive
thresholding, differential operators and color matching algorithms. Animal activity and
trajectory is significant marker in research of selected pathologies (gastrointestinal diseases
in our case]. Tracking is done in the huge database of video sequences of healthy animals
(reference] and… Show more

Abstract This article deals with tracking of laboratory animals (guinea pigs] in cage. Authors
collective applied machine vision tools based on global automatic thresholding, adaptive
thresholding, differential operators and color matching algorithms. Animal activity and
trajectory is significant marker in research of selected pathologies (gastrointestinal diseases
in our case]. Tracking is done in the huge database of video sequences of healthy animals
(reference] and animals with induced gastric pain (gastric ulcer]. Show less

Ability to explore unknown environments and ability to successfully navigate in these environments are two key abilities of any truly autonomous mobile robot. In this article, an efficient implementation of mapping and navigation using only RGBD cameras is presented. The map created by Visual SLAM techniques is converted into Octree structure, and successful A*Navigation between two points in 3D is performed. The created mapping system is able to operate in real-time even on small mobile robot… Show more

Ability to explore unknown environments and ability to successfully navigate in these environments are two key abilities of any truly autonomous mobile robot. In this article, an efficient implementation of mapping and navigation using only RGBD cameras is presented. The map created by Visual SLAM techniques is converted into Octree structure, and successful A*Navigation between two points in 3D is performed. The created mapping system is able to operate in real-time even on small mobile robot and can be easily extended with new state-of-the art techniques. Proposed way of octree encoding, due to its minimum storage requirements, enables efficient distribution of map between multiple robots over the network. Presented path-planning method with its optimizations runs in sufficient time even in higher resolution maps.

Paper will be presented at conference during September 2016 Show less

This article was motivated by the doctors’ demand to make a technical support in pathologies of gastrointestinal tract research, which would be based on machine vision tools. Proposed solution should be less expensive alternative to already existing RF methods. The objective of whole experiment was to evaluate the amount of animal motion dependent on degree of pathology. In the theoretical part of the article, several methods of animal trajectory tracking are presented: two differential… Show more

This article was motivated by the doctors’ demand to make a technical support in pathologies of gastrointestinal tract research, which would be based on machine vision tools. Proposed solution should be less expensive alternative to already existing RF methods. The objective of whole experiment was to evaluate the amount of animal motion dependent on degree of pathology. In the theoretical part of the article, several methods of animal trajectory tracking are presented: two differential methods based on background subtraction, the thresholding methods based on global and local threshold and the last method used for animal tracking was the color matching with a chosen template containing a searched spectrum of colors. The methods were tested offline on five video samples. Each sample contained situation with moving guinea pig locked in a cage under various lighting conditions.
Show less

Aim of this article is to design and to implement autonomous multi-agent system, which is used for effective mapping of unknown environment. Proposed system is compared with single mapping robot. Robots used only simple systems such as odometry and gyro-odometry for their localization. Obstacles are detected by bumpers and IR sensors. System is controlled by superior server and each agent communicates with server by TCP/IP protocol. Activity of each agent is coordinated by server and the… Show more

Aim of this article is to design and to implement autonomous multi-agent system, which is used for effective mapping of unknown environment. Proposed system is compared with single mapping robot. Robots used only simple systems such as odometry and gyro-odometry for their localization. Obstacles are detected by bumpers and IR sensors. System is controlled by superior server and each agent communicates with server by TCP/IP protocol. Activity of each agent is coordinated by server and the generated map is stored in MySQL database. Navigation of each agent is implemented in two levels - local level (obstacle avoidance) and global level (environment examination). System was verified with real robots - several iRobot Create platforms enhanced by own modules. At the end of article, the results prove that even a simple system can be effectively used for environment mapping with relatively good accuracy. © 2016 IEEE. Show less

Abstract—This article deals with tracking of laboratory animals (guinea pigs) in cage.
Authors collective applied machine vision tools based on global automatic thresholding,
adaptive thresholding, differential operators and color matching algorithms. Animal activity
and trajectory is significant marker in research of selected pathologies (gastrointestinal
diseases in our case). Tracking is done in the huge database of videosequences of healthy
animals (reference) and… Show more

Abstract—This article deals with tracking of laboratory animals (guinea pigs) in cage.
Authors collective applied machine vision tools based on global automatic thresholding,
adaptive thresholding, differential operators and color matching algorithms. Animal activity
and trajectory is significant marker in research of selected pathologies (gastrointestinal
diseases in our case). Tracking is done in the huge database of videosequences of healthy
animals (reference) and animals with induced gastric pain (gastric ulcer). The solution ... Show less

Dôležitým prvkom naozaj autonómneho mobilného robotického systému je jeho schopnosť vnímať prostredie, v ktorom sa nachádza a pohybovať sa v ňom. Aby bol robot schopný prechádzať prostredím, je výhodné, aby disponoval mapou prostredia. Mapu prostredia je možné vytvárať na základe dát zo širokého spektra snímačov. V posledných rokoch výrazne získavajú na popularite vizuálne mapovacie algoritmy používajúce RGBD kamery. Hlavným dôvodom je nízka cena takýchto zariadení. Od uvedenia lacných… Show more

Dôležitým prvkom naozaj autonómneho mobilného robotického systému je jeho schopnosť vnímať prostredie, v ktorom sa nachádza a pohybovať sa v ňom. Aby bol robot schopný prechádzať prostredím, je výhodné, aby disponoval mapou prostredia. Mapu prostredia je možné vytvárať na základe dát zo širokého spektra snímačov. V posledných rokoch výrazne získavajú na popularite vizuálne mapovacie algoritmy používajúce RGBD kamery. Hlavným dôvodom je nízka cena takýchto zariadení. Od uvedenia lacných zariadení ako napríklad snímača Microsoft Kinect nám boli predstavené viaceré techniky, ktoré dokážu vytvárať snímky priestoru pomocou tohoto snímača, čo láka robotickú komunitu ku využívaniu týchto techník pre vytvorenie mapovacích algoritmov. Problém robotického mapovania nie je však novým problémom a pri návrhu naozaj úspešných nových riešení by sme mali poznať aj histrický vývoj tejto problematiky.
Práca čitateľovi predstavuje problém vizuálneho robotického mapovania v širšom kontexte:
- problém mapovania
- prehľad aktuálne dostupných priestorových mapových reprezentácii
- problém lokalizácie
- snímače pre vizuálne mapovanie
- pravdepodobnostné metódy mapovania a lokalizácie
- techniky vizuálneho mapovania za pomoci RGBD kamier Show less

It is relatively easy to equip mobile robot with Kinect sensor, but implementation of mapping algorithm with this sensor can be complicated. In this article, we describe Kinect sensor device and two main approaches to 3D reconstruction, problem closely related to SLAM mapping problem. We experimentally evaluate reference implementations of both algorithms, compare their results and discuss published and possible improvements of both approaches.

The aim of the article is a design, execution and examination of the computer vision systems, which processes digital video, reduces noise to a minimal level, and identifies a moving object together with estimation of its distance from the camera. For the image processing, library OpenCV was used. Two different methods were examined and implemented in control system. Some results are very similar in character and functionality with the use of security camera system, but the determining the… Show more

The aim of the article is a design, execution and examination of the computer vision systems, which processes digital video, reduces noise to a minimal level, and identifies a moving object together with estimation of its distance from the camera. For the image processing, library OpenCV was used. Two different methods were examined and implemented in control system. Some results are very similar in character and functionality with the use of security camera system, but the determining the distance of a given object is a new advanced ability of proposed system. Show less