| 摘要: | 研究期間：10110~10209;It is future tendency of human being that the human life is progressively extended; especially in the advanced countries, the ratio of older-age population has increased year after year. How to improve the quality of life of seniors has become an important research topic in recent years. The wheelchair is an important transportation vehicle to help some seniors and lower body disabilities; electric wheelchair is more important autonomous vehicle for above persons without needing other help. The electric wheelchair has been developed for many years. The purpose of this research project is just to promote the autonomy and safety of the electric wheelchair based on the computer vision techniques. In this research project, we just combine the techniques of computer vision, image processing, and pattern recognition to develop the visual detection and recognition techniques for achieving the automatic following navigation purpose. The tasks in this project include: 1.vibration compensation for moving camera, 2.target and obstacle detection, 2.target recognition, and 4.target tracking. The properties of this technique include: i. the detector is not influenced by the neighboring moving objects and inserted object between the wheelchair and the guide with different appearance. ii. the wheelchair is always following the guide path to keep a fixed distance. iii. the detector possesses the collision warning function. iv. the detector is suitable for day, night, and little rainy situations. v. the detector is suitable for indoor and outdoor smooth places. vi. the system has the delay functions for moving and changing direct to improve the friendly usage. The key tasks of the safety detection techniques are accuracy and stability. The detection methods are easily influenced by weather conditions and environment factors; thus we need to spend much effort on whether the proposed methods are not influenced by various weather conditions ? whether the proposed methods may keep well performance in variant road conditions ? These considerations are characteristics of our development of the computer vision techniques for electric wheelchairs. The principal investigator of this project is an original researcher on computer vision, he has studied computer vision techniques more than twenty years; moreover, he has the application experience of computer vision for advanced safety vehicles more than ten years. From 2008, he has been a faculty in the Intelligent Mobility Technology Division, Mechanical and Systems Research Lab., ITRI to help the development of the vision detection techniques for vehicles. He has also gotten and applied the US, Taiwan, and China patents in these few years. Partial techniques are practiced and have been employed by several companies; thus we have ability to complete the execution of the research project. This is a two-year research project; in the first year, we execute: 1.vibration compensation of moving camera, 2.target detection, 3.still obstacle detection, and 4.dynamic obstacle detection. In the second year, we execute: 1.target recognition, 2.target tracking, 3.stop-and-go detection, and 4.orientation detection for different distance. Moreover, the system will be implemented on embedded systems. If the related market is mature, we can quickly combine the related companies to develop the products. Like our previous topic “visual detection and recognition techniques for advanced safety vehicles”; in this topic, we currently combine a company to develop the related products. We hope that the products would be on market after two years. |
