Description: The understanding of road information based on visual sensor is an important research direction of autonomous navigation of mobile robot, and the correct segmentation of road image
Is the key to extract effective path information. In this paper, the traditional method is difficult to realize the correct segmentation of road images in the outdoor environment with complex and disturbing factors
A method of road image segmentation based on () neural network. The method is based on the method of selecting the normalized color component of the road image
() learning algorithm neural network classifier for road and non-road identification; In order to solve the influence of ambient noise on neural network output, this paper designs serial cascade
A four-order morphological filter is used to filter the segmentation image of neural network output. The effectiveness of this method is verified by the segmentation of the measured image
And robustness, can be used in the outdoor environment of the robot's real-time visual navigation control.
To Search:
File list (Check if you may need any files):
Download