Scientific Grant Agency of the Ministry of Education, science, research and sport of the Slovak Republic and the Slovak Academy of Sciences – Grant VEGA 1/0625/14.
Visual object class recognition in video sequences using a linkage of information derived by a semantic local segmentation and a global segmentation of visual saliency.
Visual class objects recognition is one of the biggest challenges of current research in the field of computer vision. This project aims to explore new methods of recognizing classes of objects in video sequences. In the center of current research in the field of computer vision. This project aims to explore new methods of recognizing classes of objects in video sequences. In the center of research, the focus will be the research of new methods of semantic segmentation at the local level approach and segmentation of the visual saliency at the global level. An integrating part of the project proposal will be research of intelligent methods of transfer of information, which will be obtained by the local and global approach using the principle of cooperative agents.
Cultural and Educational Grant Agency MŠVVaŠ SR (KEGA) : Integration of visual information studies and creation of comprehensive multi-medial study materials.
Visual detection and object recognition are the main challenges of computer vision. Detection and recognition systems can be used in surveillance, medicine, robotics or augmented reality applications. Image pre-processing is used for removing unwanted image effects like signal noise; edge detection, blur removal, etc. in order to get images suitable for further processing.
Next step in the pipeline is the extraction of relevant and discriminative features for classification. The role of the classification is to assign the object to the correct class of objects. The last part of the book deals with color theory describing colorimetric functions and color spaces for computer vision. Lastly, visual perception theory, including eye tracking and visual saliency detection methods is described.
The book is intended not only for students, but also for the broad technical audience.
Patrik POLATSEK*, Martin PETLUŠ*, Jakub MERCZ* Lukáš SEKERÁK*, Peter HAMAR*, Róbert SABOL† Slovak University of Technology in Bratislava Faculty of Informatics and Information Technologies Ilkovičova 2, 842 16 Bratislava, Slovakia firstname.lastname@example.org
Entertainment and information systems become the important part of our lives. One of the most modern and natural human-computer types of interaction is an augmented reality (AR), where a real-world environment is supplemented by virtual data, such as visual, textual and audio data.
Our aim is to create a prototype of an interactive system with a user-friendly interface for fellow travellers in a car designed for entertainment and educational purposes. The proposed system called Carlos changes a car side window into a transparent projection screen to supply the surrounding reality with the virtual information. The system creates an AR on a car window, due to which it can inform travellers about the immediate environment in real-time.
Carlos visualises the information on a side window with a transparent film using a small LED projector. The whole system is controlled using a mobile phone with gestures, voice commands or rotation of the device.
Carlos detects interesting objects such as sightseeing, restaurants and hotels on images captured by a camera mounted on a car. In the detection phase, it compares the images using the actual GPS position and the internal database of objects of interest. The detection starts with the selection of potential objects close to the GPS position automatically received from a mobile phone. Subsequently, the object detection is performed using the feature extraction and matching methods. After the successful detection, the system computes the location of objects using the homography . Due to proper displaying the information, Carlos works with Kinect device to detect the user’s head position. Then the location of detected objects is recomputed in order to the precise placing of virtual information for the actual user’s gaze at the window. Finally, Carlos projects on a window the basic tourist textual and visual information for detected objects.
Carlos is not only the information system, but also the entertainment system. It uses the object detection also for an educational game based on the answering a question related to the object. Another AR game is a flight game which aim is to keep a plane above the horizon as long as possible. In order to detect the horizon the system detects sky regions with an edge detection algorithm. The flight of a plane is controlled by simple gestures on a screen or rotation of a smartphone.
Most AR car systems create an AR on a front window to display navigation information or increase the safety by detecting objects close to the car. An example of an entertainment system which aim is closer to our system is a project called Touch the Train Window by
* Master study programme in field: Information Systems
† Master study programme in field: Software Engineering
Supervisor: Dr. Vanda Benešová, Institute of Informatics and Software Engineering, Faculty of Informatics and Information Technologies STU in Bratislava