Offline Candidate Hand Gesture Selection and Trajectory Determination for Continuous Ethiopian Sign Language
| dc.contributor.advisor | Raimond, Kumudha (PhD) | |
| dc.contributor.author | Tsegay, Abadi | |
| dc.date.accessioned | 2018-06-27T06:21:17Z | |
| dc.date.accessioned | 2023-11-04T15:14:53Z | |
| dc.date.available | 2018-06-27T06:21:17Z | |
| dc.date.available | 2023-11-04T15:14:53Z | |
| dc.date.issued | 2011-10 | |
| dc.description.abstract | There is a clear communication gap between the deaf and hearing community. To bridge this gap, one possible solution is to teach the hearing community to use sign languages. However, a better solution is to develop a translation system that converts a continuous sign language gestures to text or speech. A lot of effort has been invested in developing alphabet recognition and continuous sign language translation systems for many sign languages around the world. In this regard, little attention has been given to Ethiopian sign language (EthSL). However, an Ethiopian Manual Alphabet (EMA) recognition system has been developed recently. For a recognition system that can recognize continuous gestures from video which can be used as a translation, a methodology that extracts candidate gestures from sequence of video frames and determines hand movement trajectories is required. In this thesis, a system that extracts candidate gestures for EMA and determines hand movement trajectories is proposed. The system has two separate parts namely Candidate Gesture Selection (CGS) and Hand Movement Trajectory Determination (HMTD). The CGS combines two metrics namely speed profile of continuous gestures and Modified Hausdorff Distance (MHD) measure and has an accuracy of 80.72%. The HMTD is done by considering each hand gesture centroid from frame to frame and using angle, x- and y-directions. A qualitative evaluation of the CGS in a correctly divided video clip is found to be 94.81%. The HMTD has an accuracy of 88.31%. The overall system performance is 71.88% Keywords: EMA, candidate gesture selection, CGS, trajectory determination, HMTD, Modified Hausdorff distance, MHD, Speed profile, search window.There is a clear communication gap between the deaf and hearing community. To bridge this gap, one possible solution is to teach the hearing community to use sign languages. However, a better solution is to develop a translation system that converts a continuous sign language gestures to text or speech. A lot of effort has been invested in developing alphabet recognition and continuous sign language translation systems for many sign languages around the world. In this regard, little attention has been given to Ethiopian sign language (EthSL). However, an Ethiopian Manual Alphabet (EMA) recognition system has been developed recently. For a recognition system that can recognize continuous gestures from video which can be used as a translation, a methodology that extracts candidate gestures from sequence of video frames and determines hand movement trajectories is required. In this thesis, a system that extracts candidate gestures for EMA and determines hand movement trajectories is proposed. The system has two separate parts namely Candidate Gesture Selection (CGS) and Hand Movement Trajectory Determination (HMTD). The CGS combines two metrics namely speed profile of continuous gestures and Modified Hausdorff Distance (MHD) measure and has an accuracy of 80.72%. The HMTD is done by considering each hand gesture centroid from frame to frame and using angle, x- and y-directions. A qualitative evaluation of the CGS in a correctly divided video clip is found to be 94.81%. The HMTD has an accuracy of 88.31%. The overall system performance is 71.88% Keywords: EMA, candidate gesture selection, CGS, trajectory determination, HMTD, Modified Hausdorff distance, MHD, Speed profile, search window.There is a clear communication gap between the deaf and hearing community. To bridge this gap, one possible solution is to teach the hearing community to use sign languages. However, a better solution is to develop a translation system that converts a continuous sign language gestures to text or speech. A lot of effort has been invested in developing alphabet recognition and continuous sign language translation systems for many sign languages around the world. In this regard, little attention has been given to Ethiopian sign language (EthSL). However, an Ethiopian Manual Alphabet (EMA) recognition system has been developed recently. For a recognition system that can recognize continuous gestures from video which can be used as a translation, a methodology that extracts candidate gestures from sequence of video frames and determines hand movement trajectories is required. In this thesis, a system that extracts candidate gestures for EMA and determines hand movement trajectories is proposed. The system has two separate parts namely Candidate Gesture Selection (CGS) and Hand Movement Trajectory Determination (HMTD). The CGS combines two metrics namely speed profile of continuous gestures and Modified Hausdorff Distance (MHD) measure and has an accuracy of 80.72%. The HMTD is done by considering each hand gesture centroid from frame to frame and using angle, x- and y-directions. A qualitative evaluation of the CGS in a correctly divided video clip is found to be 94.81%. The HMTD has an accuracy of 88.31%. The overall system performance is 71.88% Keywords: EMA, candidate gesture selection, CGS, trajectory determination, HMTD, Modified Hausdorff distance, MHD, Speed profile, search window.There is a clear communication gap between the deaf and hearing community. To bridge this gap, one possible solution is to teach the hearing community to use sign languages. However, a better solution is to develop a translation system that converts a continuous sign language gestures to text or speech. A lot of effort has been invested in developing alphabet recognition and continuous sign language translation systems for many sign languages around the world. In this regard, little attention has been given to Ethiopian sign language (EthSL). However, an Ethiopian Manual Alphabet (EMA) recognition system has been developed recently. For a recognition system that can recognize continuous gestures from video which can be used as a translation, a methodology that extracts candidate gestures from sequence of video frames and determines hand movement trajectories is required. In this thesis, a system that extracts candidate gestures for EMA and determines hand movement trajectories is proposed. The system has two separate parts namely Candidate Gesture Selection (CGS) and Hand Movement Trajectory Determination (HMTD). The CGS combines two metrics namely speed profile of continuous gestures and Modified Hausdorff Distance (MHD) measure and has an accuracy of 80.72%. The HMTD is done by considering each hand gesture centroid from frame to frame and using angle, x- and y-directions. A qualitative evaluation of the CGS in a correctly divided video clip is found to be 94.81%. The HMTD has an accuracy of 88.31%. The overall system performance is 71.88% Keywords: EMA, candidate gesture selection, CGS, trajectory determination, HMTD, Modified Hausdorff distance, MHD, Speed profile, search window. | en_US |
| dc.identifier.uri | http://etd.aau.edu.et/handle/123456789/3837 | |
| dc.language.iso | en | en_US |
| dc.publisher | Addis Ababa University | en_US |
| dc.subject | Candidate gesture selection | en_US |
| dc.subject | Cgs | en_US |
| dc.subject | Trajectory determination | en_US |
| dc.subject | Hmtd | en_US |
| dc.subject | Modified Hausdorff distance | en_US |
| dc.subject | Mhd | en_US |
| dc.subject | Speed profile | en_US |
| dc.subject | Search window | en_US |
| dc.title | Offline Candidate Hand Gesture Selection and Trajectory Determination for Continuous Ethiopian Sign Language | en_US |
| dc.type | Thesis | en_US |