Design And Implementation Of Intelligent Speech Guidance System Based On Raspberry Pi

In recent years,with the general improvement of the living standards of the people,the quality of life of disabled people has become more and more concerned by society.To this end,an increasing number of assistive devices for persons with disabilities continue to emerge,with the aim of providing them with more diverse and practical products and services for the disabled.However,in the survey,it was found that the market for blind aids in China is more chaotic,with a wide variety of products and uneven quality,and a relatively simple functional design,which is inconsistent with the current development of China’s electronics industry.Therefore,in order to serve the blind users more conveniently,it is necessary to research an intelligent voice guidance system that can assist blind people to travel safely,integrate positioning navigation,obstacle avoidance,dynamic route planning and voice interaction.The main thesis of the voice-guided system is to use embedded technology,based on the Raspberry Pi development board,while providing a variety of functions for the blind travel,including basic obstacle avoidance,positioning and navigation functions,and more advanced dynamic route planning based on turning points,combined with voice output.When the user encounters obstacles or deviates from the correct route,the system makes timely voice prompts,making the system more human and intelligent.The thesis of the guided system makes full use of the hardware conditions of the Raspberry Pi development board,and integrates modules such as voice,positioning,navigation and obstacle avoidance,so that the system usage rate is higher.In the course of the route planning of the navigation module,a navigation algorithm based on turning points is proposed in this thesis.The turning point-based navigation algorithm is a major innovation of this system.Considering the specificity of the blind group,the best way to guide blind people to walk is to use simple voice prompts to walk them instead of directly telling them the route.The turning point route navigation needs to extract the track information saved by the system,and extract a small number of turning points in the track by the turning point extraction algorithm.At this time,the dynamic route planning based on the turning point of the route is transformed into the shortest path problem of the weighted graph.In the algorithm implementation,firstly,the starting point is located in the same connected component of the graph,and then the Dijkstra algorithm is used to realize the dynamic route planning for the weighted graph.Finally,in order to ensure real-time performance,the navigation stored data is stored in the Redis database for fast reading.The advantages of this navigation method not only make navigation more accurate,but also get rid of the various constraints that must be used in the past based on traditional electronic map navigation,and make full use of the GPS acquisition data.This thesis adopts the form of voice as input and this is also an innovative point of this system.With the addition of voice input,the user can realize various functions of the guide system through simple commands of voice,and can express the place he wants to go by voice,which greatly facilitates the user and has strong practicability.