Research on Indoor Navigation System of Vision-Based Mobile Robot
|Course||Applied Computer Technology|
|Keywords||Mobile Robots Visual navigation Navigation Line Exhibition station identification Audio Playback|
With the development of modern information technology, robotics technology has been rapid advances in scientific exploration, wrecker disaster relief, military, industrial manufacturing and other fields has been widely used, at the same time, people's living standards are rising, Robotics Application examples in the entertainment industry is also increasing. Vision-based mobile robot navigation technology is the modern mobile robot navigation hotspot in recent years has been widely applied to robot navigation, where the navigation path detection problem is the visual navigation key, which directly affects the accuracy of the robot movement. Visual navigation methods to obtain the amount of information in accordance with its integrity and intelligence advantages, Robot Navigation in the future there will be a broad space for development. People engaged in a variety of work, and some work with heavy, repetitive, harmful to health and even dangerous, etc., so the need for a device to replace the people to engage in this kind of work, the robot has the mobility, flexibility and autonomy ability, and tireless robots that can adapt to various changes in the environment, so more and more widely. The system is designed for the service industry mobile robot exhibition system can be applied to galleries, museums and other exhibition activities of that nature have a place, instead of the narrator's exhibition commentary repetitive work. This paper studies the algorithms for robot navigation line, exhibition stand features logo recognition algorithms and audio reading and playing algorithms. This article discusses the research background of robot navigation, historical significance, discusses in detail the research situation, the robot navigation compared profound study pointed out various navigation methods applicable conditions, also introduced the purpose and significance of the topic . Studies in the visual navigation system used in image processing knowledge and algorithms, based on the characteristics and requirements of the system, select the image preprocessing algorithm, denoising algorithms, segmentation and color space model, the system processed images are 24 color images, in which median filtering denoising algorithm selection and combination of mathematical morphology algorithm to compare a variety of segmentation algorithms and ultimately chose the maximum between-class variance method (OTSU method) for automatic threshold segmentation and Sobel edge detection. Studied lines for robot navigation algorithm, a detailed analysis of the traditional Hough transform methods and the advantages and disadvantages of the least square method, combining this navigation road environment, this paper proposes a dual edge-based centroid, the central idea of ??the method is divided according to the edges of the navigation path, which point is calculated, respectively, of the midpoint of the upper and lower half-set of the image centroid demand, according to the obtained two directions can be calculated centroid line. The method is simple and fast algorithm for real good. Logo design and the study identified several methods, introduced a digital identification, radio frequency identification (RFID) and color identification methods, and introduces the advantages and disadvantages of these methods, combined with the indoor exhibition environment, using the HSI color space color recognition method for identification of the exhibition stand, exhibition not only meet the needs, and identify relatively simple algorithm is easy to implement. Studied several commonly used audio formats and audio playback methods, and were introduced several audio scope and the advantages and disadvantages of several ways to play audio, combined with the characteristics of the system and people's habits, and ultimately chose MCI method plays WAVE format audio files. This system is based on the DirectX mobile robot visual navigation exhibition system, configured by VC 6.0 DirectX 9.0 build experimental platform that combines indoor exhibition environment by collecting road image for image processing, segmentation of the exhibition stand (booth) and roads, identify the exhibition stand features logo, play the corresponding exhibition information, complete the exhibition commentary task, the final validation of the feasibility and system integrity.