| In the past years, the mobile robot vision navigation has become a hot field, the self-localization, environment perception and path tracking have become its key technologies, the article using the specific model-intelligence vehicle as the study object, has studied the following path navigation and the avoiding obstacle navigation with the multi-sensor information fusion,In the thesis, the main contents are as follows:On introduced the development history of robot and the characteristic of mobile robot basis, for the following path environment, a method based on the Wavelet and Mathematical Morphology for calculating the navigation parameters is proposed, using the camera sensor gathering environment information, implements the robot autonomous navigation. The gray image is obtained after enhancing and filtering by the Wavelet, removing impurities of the sides of the road by filter of Mathematical Morphology, some way is carried to make the image binarization and extract the central points of the guide lines with the Mathematical Morphology, the weighted least square method is used to calculate the parameters, which will realize the robot to navigate, the simulation experiments verify that the approach proposed above is feasible and efficient.Under no path environment, for the realization of the robot task going from the starting point to destination point, using the vision sensor to get the environment information, grid method is used to describe the whole environment, a exploring tree method is proposed for designing a global path, in the course of going forward, the partial environment information and locate self-position are constantly perceived by the ultrasonic sensor and odometer sensor, then the partial environment information and global path information are fused by the fuzzy logic method, on the basis of the types of the environment design the avoiding method. The simulation experiments verify that the way proposed above can autonomously carry out the robot avoiding obstacles and navigation in the unknown environment.With the main platform of ARM9, an autonomous navigation control experiment system is well developed with several sensors, the system consists of drive module, navigation module, vehicle turn module, sensor module and wireless communication module. For verifying the approach proposed above is feasible and efficient, it is used on the experiment system, the results of the experiment verify the system and the algorithm can meet the autonomous navigation, reach the anticipate destination. |
