Develop the Video Surveillance Set-Top-Box for Elder Falling Detection Amd Mobile Rescue Services

Project: National Science and Technology CouncilNational Science and Technology Council Academic Grants

Project Details

Abstract

Due to the growing populations of elders, the issues of activity safety become more and more important. In the last-year project, we proposed an intelligent video-surveillance system to detect falling incidents, and designed an alarming mechanism for seeking consequent emergency rescue services. The system will detect falling incidents via the image processing and pattern recognition techniques, and also will automatically issue the alarms to care-staffs in case the incidents have been detected. In addition, we develop a streaming-based parallel computing technique such that the system can support multiple cameras and the video streams can be executed concurrently in a high speed. In this year, we will initiate another two-year project to improve the system’s compatibility, scalability, and mobility. The key designs include: (1) We will use a dual-core SoC platform to implement a set-top-box supporting falling-detection and streaming-based parallel computing techniques. In this platform, the SoC chip contains two cores: one is ARM9 MCU core and the other is TI C64x DSP core. Due to the improvement on DSP processing, the SoC chip can achieve a better quality-of services for at least four-stream parallel processing. By the implementation of this set-top-box, the falling detection services can be plugged into most existing video surveillance systems. (2) We will also connect the falling detection set-top-box to a video server to support a “mobile surveillance service”. By this service, the care-staffs can monitor each camera stream via a handheld device, e.g., the smart phone, on demands. (3) The whole surveillance system will be connected to a healthcare platform, which can track the safety for high-risk elders. In case they have falling incidents, the system will identify their names and activate proper rescue process immediately. By realizing this project, the crews can obtain the following trainings: First, learn how to apply the IT techniques to implement an elder healthcare system. Second, learn how to apply the parallel programming technique to enhance software or hardware parallelism. Third, learn how to apply the distributed architecture to built up a large scale information system. And finally, learn the hardware/software co-design technique to design the embedded systems.

Project IDs

Project ID:PB9907-12655
External Project ID:NSC99-2221-E182-032
StatusFinished
Effective start/end date01/08/1031/07/11

Keywords

  • video surveillance system
  • image processing and pattern recognition
  • stream parallel processing
  • falling detection
  • mobile surveillance service

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.