Scholarlink

The global population is getting older, and it is necessary to develop new and creative solutions to ensure the safety, comfort, and privacy of senior citizens. Falling is one of the most significant risks for older individuals, which has led to a growing demand for reliable fall detection systems. However, the current Grid-EYE sensor systems are only able to detect human activities in small areas when using a single sensor, which limits their effectiveness in larger environments. To overcome this limitation, a system was developed consisting of a network of low-resolution sensors, enabling the identification of falls with 93.75% accuracy. In addition, the system can accurately determine whether there is more than one person in a room, achieving a 96.87% accuracy rate. It can also detect the precise position of a person within the monitored space. SVM models were employed to analyse sensor data, removing background noise and improving overall reliability.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Estimation of Human Localisation Using Low-Resolution IR Sensors

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Fabio Salice,
  • Serena Iarocci,
  • Chiara Licursi,
  • Sara Comai

摘要

The global population is getting older, and it is necessary to develop new and creative solutions to ensure the safety, comfort, and privacy of senior citizens. Falling is one of the most significant risks for older individuals, which has led to a growing demand for reliable fall detection systems. However, the current Grid-EYE sensor systems are only able to detect human activities in small areas when using a single sensor, which limits their effectiveness in larger environments. To overcome this limitation, a system was developed consisting of a network of low-resolution sensors, enabling the identification of falls with 93.75% accuracy. In addition, the system can accurately determine whether there is more than one person in a room, achieving a 96.87% accuracy rate. It can also detect the precise position of a person within the monitored space. SVM models were employed to analyse sensor data, removing background noise and improving overall reliability.