Segmentation Based On Spiking Neural Network Using Color Edge Gradient For Extraction Of Corridor Floor

Intelligent vehicle is to achieve independence for driving, requiring being able to initiatively perform overtaking, avoiding in the case of no intervention of human drivers. The visual navigation techniques enable mobile robots to automatically avoid obstacles in indoor and outdoor environments. So these techniques can also be used to help the blind men who lose sight. This paper selects the indoor environment as the walking scene for a blind navigation, dedicating to split the road and detect the boundary line between the ground and the wallIn this paper, for the purpose of obstacle avoidance for blind men in the environment of indoor corridor, a corridor ground segmentation algorithm is proposed using image processing mechanism of the human visual system combined with the existing segmentation algorithms in robot visual navigation techniques. The segmentation algorithm is based on a spiking neural network. Firstly, three color image gradient maps are generated utilizing a spiking neural network. The best gradient map is generated from three color components to extract the effective and useful image edges. Then threshold segmentation method is used to eliminate unwanted gradient to identify the boundary of floor. Combining the watershed algorithm can accurately and efficiently extract the corridor floor. In order to eradicate over-segmentation caused by local minima in the watershed segmentation algorithm, the region merging criteria is based on a similarity of entropy and the gray value to merge similar regions. Finally, the seed fillings algorithm to divide the divided region, with boundary tracking algorithm to derive the contour shape of the target area. A hundred different pieces of on-site corridor images have been taken in the image library, and the shooting positions are not fixed because it is supposed that a camera is equipped in the body of a blind person. The experimental results show that the algorithm works efficiently and the boundary of floor can be extracted accurately for corridor images with certain noise, textured or non-textured. The algorithm has the practicality and robustness for identification of ground floor in blind navigation.In this paper the edge gradient information is generated by a spiking neural network inspired by the visual system of the human brain, and then the watershed method is used to segmentation. In most building cases it is able to reach very good segmentation. Whatever both wall and floor are similar or completely different texture it can achieve a good robustness and adaptability. But it is not applicable for very poor lighting and harsh environments corridor. Based on this algorithm the further work is to measure the length and width of the corridor and provide important data for blind guide system, and then create a flexible "eye" for the blind.