الفهرس 
Chapter 1: IntroductionOnly 14 pages are availabe for public view
 |  | from | 163 |  |  |
| | | from | 163 | | |
Abstract
Assistive technologies for helping of blind people are demonstrating a rabid growth, providing important tools for supporting daily activities and to optimize the social inclusion. The majority of these technologies focus on assisting blind persons to navigate and also avoid obstacles. Other studies emphasized on providing them help to recognize the surrounding objects. In this regard, very few works however integrate both aspects (i.e., navigation and recognition). On contrary to the previously published works on blind guidance, this study contributes to the literature with an intelligent system, which comprises three agents into an overall system, namely, speech recognition agent, 3D image processing agent, and the path planning agent. The proposed system focuses on the speech, 3D image, and geographical information to plane the optimal path that guides the blind people when they are moving inside the indoor environment. In this thesis, the speech recognition agent acquires the verbal instructions from the blind person via the headset to determine the command of the user. Secondly, the object recognition agent starts to operate, which includes two stereo cameras for capturing the scene and forwarding the 3D coordinates information to the blind person or the path planning agent. Thirdly, a path planning agent based GIS was built to receive the distance to potential obstacles from the object recognition agent and calculate a safe path for the user to walk through. The path planning agent is based on a proposed algorithm that uses the genetic algorithm and A* algorithm to select the shortest path enabling the blind person to walk through without any obstacles.