A Wearable Dual Sensor-Glove System for the Recognition of Turkish Sign Language
Balkan 16th International Conference on Applied Sciences, Skopje, Makedonya, 24 Eylül - 29 Ekim 2026, ss.20-26, (Tam Metin Bildiri)
- Yayın Türü: Bildiri / Tam Metin Bildiri
- Basıldığı Şehir: Skopje
- Basıldığı Ülke: Makedonya
- Sayfa Sayıları: ss.20-26
- Ondokuz Mayıs Üniversitesi Adresli: Evet
Özet
Communication remains a major barrier for individuals with hearing and speech impairments, for whom sign language is the principal means of expression. In Türkiye, an estimated 250,000 people depend on Turkish Sign Language, yet a shortage of qualified interpreters limits their access to education, healthcare, and public services. TSL is structurally distinct from more widely studied sign languages, using a twenty-nine-letter two-handed fingerspelling alphabet and many two-handed lexical signs—a property that restricts most existing recognition systems, which are single-handed, tethered, and limited to small static vocabularies. This paper presents a wearable dual-glove system for sensor-based recognition of Turkish Sign Language. Each glove combines five flex sensors, which measure finger flexion, with an inertial measurement unit that captures hand orientation and motion, jointly encoding the hand-shape and movement information needed to distinguish static from dynamic signs. Signals are acquired by ESP32 microcontrollers and sent to a Raspberry Pi 5 for processing and conversion into text and speech. A rechargeable lithium-polymer power unit and wireless communication make the device portable and independent of a tethered computer. The resulting prototype offers a reproducible hardware and data-acquisition framework for recognizing Turkish Sign Language letters and selected words, with quantitative performance evaluation planned as the next phase.