期刊名称：Signal & Image Processing : An International Journal (SIPIJ)
印刷版ISSN：2229-3922
电子版ISSN：0976-710X
出版年度：2012
卷号：3
期号：3
页码：73
出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：This paper presents a new and improved concept for segmenting gestures of sign language. The algorithmpresented extracts signs from video sequences under various non static backgrounds. The signs aresegmented which are normally hands and head of the signing person by minimizing the energy function ofthe level set fused by various image characteristics such as colour, texture, boundary and shapeinformation. From RGB color video three color planes are extracted and one color plane is used based onthe contrasting environments presented by the video background. Texture edge map provides spatialinformation which makes the color features more distinctive for video segmentation. The boundary featuresare extracted by forming image edge map form the existing color and texture features. The shape of thesign is calculated dynamically and is made adaptive to each video frame for segmentation of occludeobjects. The energy minimization is achieved using level sets. Experiments show that our approachprovides excellent segmentation on signer videos for different signs under robust environments such asdiverse backgrounds, sundry illumination and different signers.
