LOGIN
>> Home
>> Topics
>> Students
>> Partners
>> Statistics


Information for topics

Topic Id:
ID topic: 582
Partner Email: dtdim@iinf.bas.bg
Project Title: Identification of a Moving Palm and Image Recognition in Dactyl Alphabet
Abstract: A team at the Institute of Information and Communication Technologies of Bulgarian Academy of Sciences (BAS) is working on an experimental system EFIRS for efficient (rapid and noise resistant) access to data bases of images (DBI), mainly following the concept CBIR (Content Based Image Retrieval). Generally said, EFIRS system extracts from a given input image the relevant content, with respect to which it organizes rapid index search for similar images in DBI. At this stage the system works with half-tone images that are well organized after the scheme “object-background”, tolerating a “moderate” level of the noise in the images. The target of EFIRS was the practice of the Patent Department of the Republic of Bulgaria, and more precisely – their large DBI of professional trademarks and services marks. At the same time, there is considerable interest towards other appropriate applications of the access methods developed in EFIRS, for example, the application mentioned in the topic of this diploma work. The dactyl (sign) alphabet is a simplified variant of the so called “sign and/or cued languages” – language versions of the communication among people with hearing loss problems. The sign alphabet (for a given language, e.g. Bulgarian), represents a set of specific positions of the palm and fingers, each one corresponding to a given letter in the written alphabet (Cyrillic, in the case). This prompts the idea of recognition – to keep a sufficient number of images for each palm sign into DBI, i.e., from a sufficient number of viewpoints, which will be later grouped with the help of EFIRS in 30 classes (each one assigned to a Cyrillic letter). Obviously, there are no limitations to the concrete alphabet considered. Purpose of the diploma work offered: To develop and study a method and a software program for real time processing of video clips of a palm, gesticulating in dactyl alphabet, for the purpose of recognition of the respective dactyl letters and words composed of them, applying the tools of EFIRS system. The development includes the following parts: - A module for extracting separate frames from a video clip; - A module for segmenting a palm in a frame and/or series of frames observing the principle that the objective target, the palm, is dominating (i.e., the largest spot) in the image of the current frame and most probably – mobile in the frames series; - A module for efficient presenting of the object (an articulating palm) for the purposes of fast search in DBI, using any of the methods for access of EFIRS system. To organize a test DBI for experiments realization, for example about 20 viewpoints for every dactyl letter, including 2-3 participants (i.e., including 20-30 specific palm shapes). Recommended environment for the software experiments – Borland C/C++ Builder.
Advisor: Assoc. Prof. Dr Dimo T. Dimov
Link:
Degree: Master
 Keywords:
Artificial intelligence & Neural networks
Computer vision
Image processing