AR Sign Language Translation via Classifier Handshape Recognition

The patent introduces graphical representations as an intermediary medium between sign language classifier constructions and spoken language. These graphical representations serve as a visual bridge that preserves the spatial and motion information from sign language while making it accessible to non-signers, thereby resolving the contradiction between communication accessibility and information preservation.

The patent transitions the translation from a one-dimensional spoken language output to a two-dimensional or three-dimensional graphical representation. This dimensional change allows the preservation of spatial path, position, and movement information that would otherwise be lost in traditional spoken language translation, enabling non-signers to perceive the full semantic content.

The patent creates visual copies or representations of the classifier constructions directly from the sign language gestures. By copying the essential visual and spatial characteristics of the handshapes and movements into graphical forms, the system maintains high translation accuracy while keeping the process relatively simple, as the graphical output directly mirrors the input gesture semantics.

The patent replaces traditional mechanical or algorithmic translation approaches with a visual representation system. Instead of converting sign language into spoken words through complex linguistic processing, the system directly generates graphical representations that preserve the motion and spatial information, simplifying the translation mechanism while improving accuracy.