Mexican sign language segmentation using color based neuronal networks to detect the individual skin color
作者：
Highlights：
• New method to detect and translate manual gestures of MLS
• The Skin color detection proposed prevents poor segmentation due to bright changes.
• The presented vision-based sign language detection avoid the use of extra devices like sensors.
• Geometric features can represent the form of the hand shape during the gestures.
• Using some representative frames of the sign we reduce the data-set.
摘要
•New method to detect and translate manual gestures of MLS•The Skin color detection proposed prevents poor segmentation due to bright changes.•The presented vision-based sign language detection avoid the use of extra devices like sensors.•Geometric features can represent the form of the hand shape during the gestures.•Using some representative frames of the sign we reduce the data-set.
论文关键词：Mexican sign language,Classification,Segmentation
论文评审过程：Received 18 January 2020, Revised 8 September 2020, Accepted 25 May 2021, Available online 6 June 2021, Version of Record 11 June 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.115295
