Research On Algorithm Based On Blind Track Detection And Direction Recognition

As an integral part of urban infrastructure construction,blind track is of great significance.The traditional travel of the visually impaired group is mostly based on the blind track,which uses the sense of feet to judge the road conditions ahead and turns.This travel method is not only very dangerous,but also cannot make full use of the blind track information.With the help of digital image processing technology,based on the blind track information,this paper takes the blind track image under external conditions such as illumination and shadow as the algorithm processing object.Starting from two aspects of blind track detection and direction recognition,different processing algorithms are proposed respectively.In blind track detection,the research work is divided into two parts: color segmentation and texture segmentation.Aiming at the color feature blind channel,a segmentation algorithm based on HSI color space and improved OTSU is proposed.The color space conversion,improved OTSU and morphological processing algorithms are used to extract image color components,achieve threshold segmentation,improve the target area,and improve Segmentation accuracy;for blind channels with obvious texture characteristics,a segmentation algorithm is proposed based on GLCM and an improved K-means algorithm.The algorithm adopts histogram equalization,gray-level co-occurrence matrix and improved K-means clustering Algorithms and the like.As the result,the contrast of the original blind channel image is enhanced,the blind channel feature value is calculated,and the position as the initial point is determined.Moreover,region segmentation is achieved,segmentation time is reduced,and the similarity between elements is enhanced.In terms of direction recognition,the processed blind track images are divided into straight blind track images and turning blind track images.For straight blind roads,the Canny edge detection operator is used to extract the boundary,then the two longest straight lines in the image are found through Hough transform,and the position of the blind road relative to the blind person is determined according to the angle between the two straight lines and the x-axis projection.It is convenient for the blind to judge the deviation direction and plan the walking route;for turning blind roads,an inflection point detection algorithm based on the polygon approximation idea is proposed,which connects the detected boundaries into a closed contour,and then uses the polygon approximation idea and the Dp algorithm to use the geometry of the contour curve.The graphics are fitted.Since the inflection points of the fitted graphics are accurate,the Susan corner detection algorithm can quickly and efficiently detect the inflection points of the blind track,avoiding the generation of false inflection points and achieving the identification of the blind track direction and improving the universality of the algorithm.Experimental results show that the algorithm proposed in this paper can effectively avoid the influence of external conditions such as illumination,improve the accuracy of segmentation,extract the blind track rangeaccurately and efficiently,and identify the deviation position of the blind track and the direction of the blind track,so as to guide the blind to travel safely and efficiently.